arXiv Daily Digest - 2026-03-17

Split Computing enables collaborative inference between edge devices and the cloud by partitioning a deep neural network into an edge-side head and a server-side tail, reducing latency and limiting exposure of raw input data. However, inference performance often degrades in practical deployments due to user-specific data distribution shifts, unreliable communication, and privacy-oriented perturbations, especially in closed environments where model architectures and parameters are inaccessible. To address this challenge, we propose SALT (Split-Adaptive Lightweight Tuning), a lightweight adaptation framework for closed Split Computing systems. SALT introduces a compact client-side adapter that refines intermediate representations produced by a frozen head network, enabling effective model adaptation without modifying the head or tail networks or increasing communication overhead. By modifying only the training conditions, SALT supports multiple adaptation objectives, including user personalization, communication robustness, and privacy-aware inference. Experiments using ResNet-18 on CIFAR-10 and CIFAR-100 show that SALT achieves higher accuracy than conventional retraining and fine-tuning while significantly reducing training cost. On CIFAR-10, SALT improves personalized accuracy from 88.1% to 93.8% while reducing training latency by more than 60%. SALT also maintains over 90% accuracy under 75% packet loss and preserves high accuracy (about 88% at sigma = 1.0) under noise injection. These results demonstrate that SALT provides an efficient and practical adaptation framework for real-world Split Computing systems.

Spiking Federated Learning (SFL) has been widely studied with the energy efficiency of Spiking Neural Networks (SNNs). However, existing SFL methods require model homogeneity and assume all clients have sufficient computational resources, resulting in the exclusion of some resource-constrained clients. To address the prevalent system heterogeneity in real-world scenarios, enabling heterogeneous SFL systems that allow clients to adaptively deploy models of different scales based on their local resources is crucial. To this end, we introduce SFedHIFI, a novel Spiking Federated Learning framework with Fire Rate-Based Heterogeneous Information Fusion. Specifically, SFedHIFI employs channel-wise matrix decomposition to deploy SNN models of adaptive complexity on clients with heterogeneous resources. Building on this, the proposed heterogeneous information fusion module enables cross-scale aggregation among models of different widths, thereby enhancing the utilization of diverse local knowledge. Extensive experiments on three public benchmarks demonstrate that SFedHIFI can effectively enable heterogeneous SFL, consistently outperforming all three baseline methods. Compared with ANN-based FL, it achieves significant energy savings with only a marginal trade-off in accuracy.

Large multimodal models (LMMs) have recently demonstrated remarkable performance in video question answering (VideoQA), yet reasoning over video remains challenging due to high inference cost and diluted information. Keyframe selection offers efficiency and sharper reasoning but suffers from sparse supervision and redundant frame choices when relying only on image-text similarity. We present a question-aware keyframe selection framework with two components: pseudo keyframe labels derived from LMMs that provide informative supervision and a coverage regularization that promotes diverse, complementary evidence across time. Experiments on NExT-QA show that our method significantly improves accuracy, especially for temporal and causal question types, establishing keyframe selection as an effective and learnable module for VideoQA.

Machine learning models deployed in critical care settings exhibit demographic biases, particularly gender disparities, that undermine clinical trust and equitable treatment. This paper introduces FairMed-XGB, a novel framework that systematically detects and mitigates gender-based prediction bias while preserving model performance and transparency. The framework integrates a fairness-aware loss function combining Statistical Parity Difference, Theil Index, and Wasserstein Distance, jointly optimised via Bayesian Search into an XGBoost classifier. Post-mitigation evaluation on seven clinically distinct cohorts derived from the MIMIC-IV-ED and eICU databases demonstrates substantial bias reduction: Statistical Parity Difference decreases by 40 to 51 percent on MIMIC-IV-ED and 10 to 19 percent on eICU; Theil Index collapses by four to five orders of magnitude to near-zero values; Wasserstein Distance is reduced by 20 to 72 percent. These gains are achieved with negligible degradation in predictive accuracy (AUC-ROC drop <0.02). SHAP-based explainability reveals that the framework diminishes reliance on gender-proxy features, providing clinicians with actionable insights into how and where bias is corrected. FairMed-XGB offers a robust, interpretable, and ethically aligned solution for equitable clinical decision-making, paving the way for trustworthy deployment of AI in high-stakes healthcare environments.

Spiking Neural Networks (SNNs) provide energy-efficient computation but their deployment is constrained by dense connectivity and high spiking operation costs. Existing magnitude-based pruning strategies, when naively applied to SNNs, fail to account for temporal accumulation, non-uniform timestep contributions, and membrane stability, often leading to severe performance degradation. This paper proposes Spiking Layer-Adaptive Magnitude-based Pruning (SLAMP), a theory-guided pruning framework that generalizes layer-adaptive magnitude pruning to temporal SNNs by explicitly controlling worst-case output distortion across layers and timesteps. SLAMP formulates sparsity allocation as a temporal distortion-constrained optimization problem, yielding time-aware layer importance scores that reduce to conventional layer-adaptive pruning in single-timestep limit. An efficient two-stage procedure is derived, combining temporal score estimation, global sparsity allocation, and magnitude pruning with retraining for stability recovery. Experiments on CIFAR10, CIFAR100, and the event-based CIFAR10-DVS datasets demonstrate that SLAMP achieves substantial connectivity and spiking operation reductions while preserving accuracy, enabling efficient and deployable SNN inference.

Complex dynamical systems-such as climate, ecosystems, and economics-can undergo catastrophic and potentially irreversible regime changes, often triggered by environmental parameter drift and stochastic disturbances. These critical thresholds, known as tipping points, pose a prediction problem of both theoretical and practical significance, yet remain largely unresolved. To address this, we articulate a model-free framework that integrates the measures characterizing the stability and sensitivity of dynamical systems with the reservoir computing (RC), a lightweight machine learning technique, using only observational time series data. The framework consists of two stages. The first stage involves using RC to robustly learn local complex dynamics from observational data segmented into windows. The second stage focuses on accurately detecting early warning signals of tipping points by analyzing the learned autonomous RC dynamics through dynamical measures, including the dominant eigenvalue of the Jacobian matrix, the maximum Floquet multiplier, and the maximum Lyapunov exponent. Furthermore, when these dynamical measures exhibit trend-like patterns, their extrapolation enables ultra-early prediction of tipping points significantly prior to the occurrence of critical transitions. We conduct a rigorous theoretical analysis of the proposed method and perform extensive numerical evaluations on a series of representative synthetic systems and eight real-world datasets, as well as quantitatively predict the tipping time of the Atlantic Meridional Overturning Circulation system. Experimental results demonstrate that our framework exhibits advantages over the baselines in comprehensive evaluations, particularly in terms of dynamical interpretability, prediction stability and robustness, and ultra-early prediction capability.

Remote sensing world models aim to both explain observed changes and forecast plausible futures, two tasks that share spatiotemporal priors. Existing methods, however, typically address them separately, limiting cross-task transfer. We present RS-WorldModel, a unified world model for remote sensing that jointly handles spatiotemporal change understanding and text-guided future scene forecasting, and we build RSWBench-1.1M, a 1.1 million sample dataset with rich language annotations covering both tasks. RS-WorldModel is trained in three stages: (1) Geo-Aware Generative Pre-training (GAGP) conditions forecasting on geographic and acquisition metadata; (2) synergistic instruction tuning (SIT) jointly trains understanding and forecasting; (3) verifiable reinforcement optimization (VRO) refines outputs with verifiable, task-specific rewards. With only 2B parameters, RS-WorldModel surpasses open-source models up to 120$ \times $ larger on most spatiotemporal change question-answering metrics. It achieves an FID of 43.13 on text-guided future scene forecasting, outperforming all open-source baselines as well as the closed-source Gemini-2.5-Flash Image (Nano Banana).

Reliable control and state estimation of differential drive robots (DDR) operating in dynamic and uncertain environments remains a challenge, particularly when system dynamics are partially unknown and sensor measurements are prone to degradation. This work introduces a unified control and state estimation framework that combines a Lyapunov-based nonlinear controller and Adaptive Neural Networks (ANN) with Extended Kalman Filter (EKF)-based multi-sensor fusion. The proposed controller leverages the universal approximation property of neural networks to model unknown nonlinearities in real time. An online adaptation scheme updates the weights of the radial basis function (RBF), the architecture chosen for the ANN. The learned dynamics are integrated into a feedback linearization (FBL) control law, for which theoretical guarantees of closed-loop stability and asymptotic convergence in a trajectory-tracking task are established through a Lyapunov-like stability analysis. To ensure robust state estimation, the EKF fuses inertial measurement unit (IMU) and odometry from monocular, 2D-LiDAR and wheel encoders. The fused state estimate drives the intelligent controller, ensuring consistent performance even under drift, wheel slip, sensor noise and failure. Gazebo simulations and real-world experiments are done using DDR, demonstrating the effectiveness of the approach in terms of improved velocity tracking performance with reduction in linear and angular velocity errors up to $53.91\%$ and $29.0\%$ in comparison to the baseline FBL.

Text-rich graphs, which integrate complex structural dependencies with abundant textual information, are ubiquitous yet remain challenging for existing learning paradigms. Conventional methods and even LLM-hybrids compress rich text into static embeddings or summaries before structural reasoning, creating an information bottleneck and detaching updates from the raw content. We argue that in text-rich graphs, the text is not merely a node attribute but the primary medium through which structural relationships are manifested. We introduce RAMP, a Raw-text Anchored Message Passing approach that moves beyond using LLMs as mere feature extractors and instead recasts the LLM itself as a graph-native aggregation operator. RAMP exploits the text-rich nature of the graph via a novel dual-representation scheme: it anchors inference on each node's raw text during each iteration while propagating dynamically optimized messages from neighbors. It further handles both discriminative and generative tasks under a single unified generative formulation. Extensive experiments show that RAMP effectively bridges the gap between graph propagation and deep text reasoning, achieving competitive performance and offering new insights into the role of LLMs as graph kernels for general-purpose graph learning.

We introduce a novel self-supervised learning framework that automatically learns representations from input computer-aided design (CAD) models for downstream tasks, including part classification, modeling segmentation, and machining feature recognition. To train our network, we construct a large-scale, unlabeled dataset of boundary representation (BRep) models. The success of our algorithm relies on two keycomponents. The first is a masked graph autoencoder that reconstructs randomly masked geometries and attributes of BReps for representation learning to enhance the generalization. The second is a hierarchical graph Transformer architecture that elegantly fuses global and local learning by a cross-scale mutual attention block to model long-range geometric dependencies and a graph neural network block to aggregate local topological information. After training the autoencoder, we replace its decoder with a task-specific network trained on a small amount of labeled data for downstream tasks. We conduct experiments on various tasks and achieve high performance, even with a small amount of labeled data, demonstrating the practicality and generalizability of our model. Compared to other methods, our model performs significantly better on downstream tasks with the same amount of training data, particularly when the training data is very limited.

We introduce directional routing, a lightweight mechanism that gives each transformer attention head learned suppression directions controlled by a shared router, at 3.9% parameter cost. We train a 433M-parameter model alongside an identical baseline in a single run, then trace the resulting circuits through mechanistic interpretability. Routing becomes the model's dominant computational pathway. Disabling it collapses factual recall to near-zero probability across all 8 test prompts and drops induction accuracy from 93.4% to 0.0%. Knocking out individual attention heads has negligible effect: the primary mover head's removal actually increases target probability, and induction heads retain 98.6% accuracy without their strongest member. The coordination mechanism is irreplaceable; the components it coordinates are not. The model also self-organizes, without explicit pressure, into two regimes: domain-adaptive routing in early layers and fixed syntactic pruning in late layers, where the least-varying layer is the most critical (+42.6 PPL when disabled). Routing reduces perplexity 31-56% relative to the baseline, though downstream multiple-choice benchmarks do not yet reflect these gains.

Model-based approaches for (bio)process systems often suffer from incomplete knowledge of the underlying physical, chemical, or biological laws. Universal differential equations, which embed neural networks within differential equations, have emerged as powerful tools to learn this missing physics from experimental data. However, neural networks are inherently opaque, motivating their post-processing via symbolic regression to obtain interpretable mathematical expressions. Genetic algorithm-based symbolic regression is a popular approach for this post-processing step, but provides only point estimates and cannot quantify the confidence we should place in a discovered equation. We address this limitation by applying Bayesian symbolic regression, which uses Reversible Jump Markov Chain Monte Carlo to sample from the posterior distribution over symbolic expression trees. This approach naturally quantifies uncertainty in the recovered model structure. We demonstrate the methodology on a Lotka-Volterra predator-prey system and then show how a well-designed experiment leads to lower uncertainty in a fed-batch bioreactor case study.

Spectrogram-based representations have grown to dominate the feature space for deep learning audio analysis systems, and are often adopted for speech analysis also. Initially, the primary motivator for spectrogram-based representations was their ability to present sound as a two dimensional signal in the time-frequency plane, which not only provides an interpretable physical basis for analysing sound, but also unlocks the use of a wide range of machine learning techniques such as convolutional neural networks, that had been developed for image processing. A spectrogram is a matrix characterised by the resolution and span of its two dimensions, as well as by the representation and scaling of each element. Many possibilities for these three characteristics have been explored by researchers across numerous application areas, with different settings showing affinity for various tasks. This paper reviews the use of spectrogram-based representations and surveys the state-of-the-art to question how front-end feature representation choice allies with back-end classifier architecture for different tasks.

We present a fine-tuned RoBERTa-base classifier (125M parameters) for mapping Common Vulnerabilities and Exposures (CVE) descriptions to Common Weakness Enumeration (CWE) categories. We construct a large-scale training dataset of 234,770 CVE descriptions with AI-refined CWE labels using Claude Sonnet 4.6, and agreement-filtered evaluation sets where NVD and AI labels agree. On our held-out test set (27,780 samples, 205 CWE classes), the model achieves 87.4% top-1 accuracy and 60.7% Macro F1 -- a +15.5 percentage-point Macro F1 gain over a TF-IDF baseline that already reaches 84.9% top-1, demonstrating the model's advantage on rare weakness categories. On the external CTI-Bench benchmark (NeurIPS 2024), the model achieves 75.6% strict accuracy (95% CI: 72.8-78.2%) -- statistically indistinguishable from Cisco Foundation-Sec-8B-Reasoning (75.3%, 8B parameters) at 64x fewer parameters. We release the dataset, model, and training code.

Geoenergy projects (CO2 storage, geothermal, subsurface H2 generation/storage, critical minerals from subsurface fluids, or nuclear waste disposal) increasingly follow a petroleum-style funnel from screening and appraisal to operations, monitoring, and stewardship. Across this funnel, limited and heterogeneous observations must be turned into risk-bounded operational choices under strong physical and geological constraints - choices that control deployment rate, cost of capital, and the credibility of climate-mitigation claims. These choices are inherently multi-objective, balancing performance against containment, pressure footprint, induced seismicity, energy/water intensity, and long-term stewardship. We argue that progress is limited by four recurring bottlenecks: (i) scarce, biased labels and few field performance outcomes; (ii) uncertainty treated as an afterthought rather than the deliverable; (iii) weak scale-bridging from pore to basin (including coupled chemical-flow-geomechanics); and (iv) insufficient quality assurance (QA), auditability, and governance for regulator-facing deployment. We outline machine learning (ML) approaches that match these realities (hybrid physics-ML, probabilistic uncertainty quantification (UQ), structure-aware representations, and multi-fidelity/continual learning) and connect them to four anchor applications: imaging-to-process digital twins, multiphase flow and near-well conformance, monitoring and inverse problems (monitoring, measurement, and verification (MMV), including deformation and microseismicity), and basin-scale portfolio management. We close with a pragmatic agenda for benchmarks, validation, reporting standards, and policy support needed for reproducible and defensible ML in sustainable geoenergy.

Large language models (LLMs) have recently demonstrated promising performance in simultaneous machine translation (SimulMT). However, applying decoder-only LLMs to SimulMT introduces a positional mismatch, which leads to a dilemma between decoding efficiency and positional consistency. Existing approaches often rely on specific positional encodings or carefully designed prompting schemes, and thus fail to simultaneously achieve inference efficiency, positional consistency, and broad model compatibility. In this work, we propose ExPosST, a general framework that resolves this dilemma through explicit position allocation. ExPosST reserves fixed positional slots for incoming source tokens, enabling efficient decoding with KV cache across different positional encoding methods. To further bridge the gap between fine-tuning and inference, we introduce a policy-consistent fine-tuning strategy that aligns training with inference-time decoding behavior. Experiments across multiple language pairs demonstrate that ExPosST effectively supports simultaneous translation under diverse policies.

In this paper, machine learning techniques are used to forecast the difference between bike returns and withdrawals at each station of a bike sharing system. The forecasts are integrated into a simulation framework that is used to support long-term decisions and model the daily dynamics, including the relocation of bikes. We assess the quality of the machine learning-based forecasts in two ways. Firstly, we compare the forecasts with alternative prediction methods. Secondly, we analyze the impact of the forecasts on the quality of the output of the simulation framework. The evaluation is based on real-world data of the bike sharing system currently operating in Brescia, Italy.

Photonic quantum processors naturally produce intrinsically stochastic measurement outcomes, offering a hardware-native source of structured randomness that can be exploited during machine-learning training. Here we introduce Photonic Quantum-Enhanced Knowledge Distillation (PQKD), a hybrid quantum photonic--classical framework in which a programmable photonic circuit generates a compact conditioning signal that constrains and guides a parameter-efficient student network during distillation from a high-capacity teacher. PQKD replaces fully trainable convolutional kernels with dictionary convolutions: each layer learns only a small set of shared spatial basis filters, while sample-dependent channel-mixing weights are derived from shot-limited photonic features and mapped through a fixed linear transform. Training alternates between standard gradient-based optimisation of the student and sampling-robust, gradient-free updates of photonic parameters, avoiding differentiation through photonic hardware. Across MNIST, Fashion-MNIST and CIFAR-10, PQKD traces a controllable compression--accuracy frontier, remaining close to teacher performance on simpler benchmarks under aggressive convolutional compression. Performance degrades predictably with finite sampling, consistent with shot-noise scaling, and exponential moving-average feature smoothing suppresses high-frequency shot-noise fluctuations, extending the practical operating regime at moderate shot budgets.

Cancer pathological analysis requires modeling tumor heterogeneity across multiple modalities, primarily through transcriptomics and whole slide imaging (WSI), along with their spatial relations. On one hand, bulk transcriptomics and WSI images are largely available but lack spatial mapping; on the other hand, spatial transcriptomics (ST) data can offer high spatial resolution, yet facing challenges of high cost, low sequencing depth, and limited sample sizes. Therefore, the data foundation of either side is flawed and has its limit in accurately finding the mapping between the two modalities. To this end, we propose BiTro, a bidirectional transfer learning framework that can enhance bulk and spatial transcriptomics prediction from pathological images. Our contributions are twofold. First, we design a universal and transferable model architecture that works for both bulk+WSI and ST data. A major highlight is that we model WSI images on the cellular level to better capture cells' visual features, morphological phenotypes, and their spatial relations; to map cells' features to their transcriptomics measured in bulk or ST, we adopt multiple instance learning. Second, by using LoRA, our model can be efficiently transferred between bulk and ST data to exploit their complementary information. To test our framework, we conducted comprehensive experiments on five cancer datasets. Results demonstrate that 1) our base model can achieve better or competitive performance compared to existing models on bulk or spatial transcriptomics prediction, and 2) transfer learning can further improve the base model's performance.

Trust and ethical concerns due to the widespread deployment of opaque machine learning (ML) models motivating the need for reliable model explanations. Post-hoc model-agnostic explanation methods addresses this challenge by learning a surrogate model that approximates the behavior of the deployed black-box ML model in the locality of a sample of interest. In post-hoc scenarios, neither the underlying model parameters nor the training are available, and hence, this local neighborhood must be constructed by generating perturbed inputs in the neighborhood of the sample of interest, and its corresponding model predictions. We propose \emph{Expected Active Gain for Local Explanations} (\texttt{EAGLE}), a post-hoc model-agnostic explanation framework that formulates perturbation selection as an information-theoretic active learning problem. By adaptively sampling perturbations that maximize the expected information gain, \texttt{EAGLE} efficiently learns a linear surrogate explainable model while producing feature importance scores along with the uncertainty/confidence estimates. Theoretically, we establish that cumulative information gain scales as $\mathcal{O}(d \log t)$, where $d$ is the feature dimension and $t$ represents the number of samples, and that the sample complexity grows linearly with $d$ and logarithmically with the confidence parameter $1/δ$. Empirical results on tabular and image datasets corroborate our theoretical findings and demonstrate that \texttt{EAGLE} improves explanation reproducibility across runs, achieves higher neighborhood stability, and improves perturbation sample quality as compared to state-of-the-art baselines such as Tilia, US-LIME, GLIME and BayesLIME.

Large language models (LLMs) are evaluated for calibration using metrics such as Expected Calibration Error that conflate two distinct components: the model's ability to discriminate correct from incorrect answers (sensitivity) and its tendency toward confident or cautious responding (bias). Signal Detection Theory (SDT) decomposes these components. While SDT-derived metrics such as AUROC are increasingly used, the full parametric framework - unequal-variance model fitting, criterion estimation, z-ROC analysis - has not been applied to LLMs as signal detectors. In this pre-registered study, we treat three LLMs as observers performing factual discrimination across 168,000 trials and test whether temperature functions as a criterion shift analogous to payoff manipulations in human psychophysics. Critically, this analogy may break down because temperature changes the generated answer itself, not only the confidence assigned to it. Our results confirm the breakdown with temperature simultaneously increasing sensitivity (AUC) and shifting criterion. All models exhibited unequal-variance evidence distributions (z-ROC slopes 0.52-0.84), with instruct models showing more extreme asymmetry (0.52-0.63) than the base model (0.77-0.87) or human recognition memory (~0.80). The SDT decomposition revealed that models occupying distinct positions in sensitivity-bias space could not be distinguished by calibration metrics alone, demonstrating that the full parametric framework provides diagnostic information unavailable from existing metrics.

Automated essay scoring (AES) predicts multiple rubric-defined trait scores for each essay, where each trait follows an ordered discrete rating scale. Most LLM-based AES methods cast scoring as autoregressive token generation and obtain the final score via decoding and parsing, making the decision implicit. This formulation is particularly sensitive in multimodal AES, where the usefulness of visual inputs varies across essays and traits. To address these limitations, we propose Decision-Level Ordinal Modeling (DLOM), which makes scoring an explicit ordinal decision by reusing the language model head to extract score-wise logits on predefined score tokens, enabling direct optimization and analysis in the score space. For multimodal AES, DLOM-GF introduces a gated fusion module that adaptively combines textual and multimodal score logits. For text-only AES, DLOM-DA adds a distance-aware regularization term to better reflect ordinal distances. Experiments on the multimodal EssayJudge dataset show that DLOM improves over a generation-based SFT baseline across scoring traits, and DLOM-GF yields further gains when modality relevance is heterogeneous. On the text-only ASAP/ASAP++ benchmarks, DLOM remains effective without visual inputs, and DLOM-DA further improves performance and outperforms strong representative baselines.

The rapid evolution of end-to-end spoken dialogue systems demands transcending mere textual semantics to incorporate paralinguistic nuances and the spontaneous nature of human conversation. However, current methods struggle with two critical gaps: the modality gap, involving prosody and emotion, and the colloquialness gap, distinguishing written scripts from natural speech. To address these challenges, we introduce SDiaReward, an end-to-end multi-turn reward model trained on SDiaReward-Dataset, a novel collection of episode-level preference pairs explicitly targeting these gaps. It operates directly on full multi-turn speech episodes and is optimized with pairwise preference supervision, enabling joint assessment of modality and colloquialness in a single evaluator. We further establish ESDR-Bench, a stratified benchmark for robust episode-level evaluation. Experiments demonstrate that SDiaReward achieves state-of-the-art pairwise preference accuracy, significantly outperforming general-purpose audio LLMs. Further analysis suggests that SDiaReward captures relative conversational expressiveness beyond superficial synthesis cues, improving generalization across domains and recording conditions. Code, data, and demos are available at https://sdiareward.github.io/.

ChatGPT, with its customization features and Voice Mode, has the potential for more engaging and peresonalized ESL (English as a Second Language) education. This study examines the efficacy of customized ChatGPT conversational features in facilitating ESL speaking practices, comparing the performance of four versions of ChatGPT Voice Mode: uncustomized Standard mode, uncustomized Advanced mode, customized Standard mode, and customized Advanced mode. Customization was guided by prompt engineering principles and grounded in relevant theories, including Motivation Theory, Culturally Responsive Teaching (CRT), Communicative Language Teaching (CLT), and the Affective Filter Hypothesis. Content analysis found that customized versions generally provided more balanced feedback and emotional support, contributing to a positive and motivating learning environment. However, cultural responsiveness did not show significant improvement despite targeted customization efforts. These initial findings suggest that customization could enhance ChatGPT's capacity as a more effective language tutor, with the standard model already capable of meeting the learning needs. The study underscores the importance of prompt engineering and AI literacy in maximizaing AI's potential in language learning.

Objectives: Full-waveform inversion (FWI) is a high-resolution geophysical imaging technique that reconstructs subsurface velocity models by iteratively minimizing the misfit between predicted and observed seismic data. However, under complex geological conditions, conventional FWI suffers from strong dependence on the initial model and tends to produce unstable results when the data are sparse or contaminated by noise. Methods: To address these limitations, this paper proposes a physics-driven generative adversarial network-based full-waveform inversion method. The proposed approach integrates the data-driven capability of deep neural networks with the physical constraints imposed by the seismic wave equation, and employs adversarial training through a discriminator to enhance the stability and robustness of the inversion results. Results: Experimental results on two representative benchmark geological models demonstrate that the proposed method can effectively recover complex velocity structures and achieves superior performance in terms of structural similarity (SSIM) and signal-to-noise ratio (SNR). Conclusions: This method provides a promising solution for alleviating the initial-model dependence in full-waveform inversion and shows strong potential for practical applications.

This research paper outlines the development and implementation of a novel Clinical Decision Support System (CDSS) that integrates AI predictive modeling with medical knowledge bases. It utilizes the quantifiable information elements in lab results for inferring likely diagnoses a patient might have. Subsequently, suggesting investigations to confirm the likely diagnoses -- an assistive tool for physicians. The system fuses knowledge contained in a rule-base expert system with inferences of data driven predictors based on the features in labs. The data for 593,055 patients was collected from 547 primary care centers across the US to model our decision support system and derive Real-Word Evidence (RWE) to make it relevant for a large demographic of patients. Our Rule-Base comprises clinically validated rules, modeling 59 health conditions that can directly confirm one or more of diseases and assign ICD-10 codes to them. The Likely Diagnosis system uses multi-class classification, covering 37 ICD-10 codes, which are grouped together into 11 categories based on the labs that physicians prescribe to confirm the diagnosis. This research offers a novel system that assists a physician by utilizing medical profile of a patient and routine lab investigations to predict a group of likely diseases and then confirm them, coupled with providing explanations for inferences, thereby assisting physicians to reduce misdiagnosis of patients in clinical decision-making.

Low-resource languages serve as invaluable repositories of human history, preserving cultural and intellectual diversity. Despite their significance, they remain largely absent from modern natural language processing systems. While progress has been made for widely spoken African languages such as Swahili, Yoruba, and Amharic, smaller indigenous languages like Efik continue to be underrepresented in machine translation research. This study evaluates the effectiveness of state-of-the-art multilingual neural machine translation models for English-Efik translation, leveraging a small-scale, community-curated parallel corpus of 13,865 sentence pairs. We fine-tuned both the mT5 multilingual model and the NLLB200 model on this dataset. NLLB-200 outperformed mT5, achieving BLEU scores of 26.64 for English-Efik and 31.21 for Efik-English, with corresponding chrF scores of 51.04 and 47.92, indicating improved fluency and semantic fidelity. Our findings demonstrate the feasibility of developing practical machine translation tools for low-resource languages and highlight the importance of inclusive data practices and culturally grounded evaluation in advancing equitable NLP.

Predicting immunoglobulin-antigen (Ig-Ag) binding remains a significant challenge due to the paucity of experimentally-resolved complexes and the limited accuracy of de novo Ig structure prediction. We introduce IgPose, a generalizable framework for Ig-Ag pose identification and scoring, built on a generative data-augmentation pipeline. To mitigate data scarcity, we constructed the Structural Immunoglobulin Decoy Database (SIDD), a comprehensive repository of high-fidelity synthetic decoys. IgPose integrates equivariant graph neural networks, ESM-2 embeddings, and gated recurrent units to synergistically capture both geometric and evolutionary features. We implemented interface-focused k-hop sampling with biologically guided pooling to enhance generalization across diverse interfaces. The framework comprises two sub-networks--IgPoseClassifier for binding pose discrimination and IgPoseScore for DockQ score estimation--and achieves robust performance on curated internal test sets and the CASP-16 benchmark compared to physics and deep learning baselines. IgPose serves as a versatile computational tool for high-throughput antibody discovery pipelines by providing accurate pose filtering and ranking. IgPose is available on GitHub (https://github.com/arontier/igpose).

Reliable photovoltaic (PV) power generation requires timely detection of module defects that may reduce energy yield, accelerate degradation, and increase lifecycle operation and maintenance costs during field operation. Electroluminescence (EL) imaging has therefore been widely adopted for PV module inspection. However, automated defect detection in real operational environments remains challenging due to heterogeneous module geometries, low-resolution imaging conditions, subtle defect morphology, long-tailed defect distributions, and continual data shifts introduced by evolving inspection and labeling processes. These factors significantly limit the robustness and long-term maintainability of conventional deep-learning inspection pipelines. To address these challenges, this paper proposes SEPDD, a Self-Evolving Photovoltaic Defect Detection framework designed for evolving industrial PV inspection scenarios. SEPDD integrates automated model optimization with a continual self-evolving learning mechanism, enabling the inspection system to progressively adapt to distribution shifts and newly emerging defect patterns during long-term deployment. Experiments conducted on both a public PV defect benchmark and a private industrial EL dataset demonstrate the effectiveness of the proposed framework. Both datasets exhibit severe class imbalance and significant domain shift. SEPDD achieves a leading mAP50 of 91.4% on the public dataset and 49.5% on the private dataset. It surpasses the autonomous baseline by 14.8% and human experts by 4.7% on the public dataset, and by 4.9% and 2.5%, respectively, on the private dataset.

Many strategic decision-making problems, such as environment design for warehouse robots, can be naturally formulated as bi-level reinforcement learning (RL), where a leader agent optimizes its objective while a follower solves a Markov decision process (MDP) conditioned on the leader's decisions. In many situations, a fundamental challenge arises when the leader cannot intervene in the follower's optimization process; it can only observe the optimization outcome. We address this decentralized setting by deriving the hypergradient of the leader's objective, i.e., the gradient of the leader's strategy that accounts for changes in the follower's optimal policy. Unlike prior hypergradient-based methods that require extensive data for repeated state visits or rely on gradient estimators whose complexity can increase substantially with the high-dimensional leader's decision space, we leverage the Boltzmann covariance trick to derive an alternative hypergradient formulation. This enables efficient hypergradient estimation solely from interaction samples, even when the leader's decision space is high-dimensional. Additionally, to our knowledge, this is the first method that enables hypergradient-based optimization for 2-player Markov games in decentralized settings. Experiments highlight the impact of hypergradient updates and demonstrate our method's effectiveness in both discrete and continuous state tasks.

In e-commerce, LLM agents show promise for shopping tasks such as recommendations, budgeting, and bundle deals, where accurately capturing user preferences from long-term conversations is critical. However, two challenges hinder realizing this potential: (1) the absence of benchmarks for evaluating long-term preference-aware shopping tasks, and (2) the lack of end-to-end optimization due to existing designs that treat preference identification and shopping assistance as separate components. In this paper, we introduce a novel benchmark with a long-term memory setup, spanning two shopping tasks over 1.2 million real-world products, and propose Shopping Companion, a unified framework that jointly tackles memory retrieval and shopping assistance while supporting user intervention. To train such capabilities, we develop a dual-reward reinforcement learning strategy with tool-wise rewards to handle the sparse and discontinuous rewards inherent in multi-turn interactions. Experimental results demonstrate that even state-of-the-art models (such as GPT-5) achieve success rates under 70% on our benchmark, highlighting the significant challenges in this domain. Notably, our lightweight LLM, trained with Shopping Companion, consistently outperforms strong baselines, achieving better preference capture and task performance, which validates the effectiveness of our unified design.

We introduce a score-filter-enhanced data assimilation framework designed to reduce predictive uncertainty in machine learning (ML) models for data-driven dynamical system forecasting. Machine learning serves as an efficient numerical model for predicting dynamical systems. However, even with sufficient data, model uncertainty remains and accumulates over time, causing the long-term performance of ML models to deteriorate. To overcome this difficulty, we integrate data assimilation techniques into the training process to iteratively refine the model predictions by incorporating observational information. Specifically, we apply the Ensemble Score Filter (EnSF), a generative AI-based training-free diffusion model approach, for solving the data assimilation problem in high-dimensional nonlinear complex systems. This leads to a hybrid data assimilation-training framework that combines ML with EnSF to improve long-term predictive performance. We shall demonstrate that EnSF-enhanced ML can effectively reduce predictive uncertainty in ML-based Lorenz-96 system prediction and the Korteweg-De Vries (KdV) equation prediction.

Urban traffic management increasingly requires intelligent sensing systems capable of adapting to dynamic traffic conditions without costly infrastructure modifications. Vision-based vehicle detection has therefore become a key technology for modern intelligent transportation systems. This study presents Video Detector (VD), a dual-phase vision-based traffic intersection management system designed as a flexible and cost-effective alternative to traditional inductive loop detectors. The framework integrates a real-time module (VD-RT) for intersection control with an offline analytical module (VD-Offline) for detailed traffic behavior analysis. Three system configurations were implemented using SSD Inception v2, Faster R-CNN Inception v2, and CenterNet ResNet-50 V1 FPN, trained on datasets totaling 108,000 annotated images across 6-10 vehicle classes. Experimental results show detection performance of up to 90% test accuracy and 29.5 mAP@0.5, while maintaining real-time throughput of 37 FPS on HD video streams. Field deployments conducted in collaboration with Istanbul IT and Smart City Technologies Inc. (ISBAK) demonstrate stable operation under diverse environmental conditions. The system supports virtual loop detection, vehicle counting, multi-object tracking, queue estimation, speed analysis, and multiclass vehicle classification, enabling comprehensive intersection monitoring without the need for embedded road sensors. The annotated dataset and training pipeline are publicly released to support reproducibility. These results indicate that the proposed framework provides a scalable and deployable vision-based solution for intelligent transportation systems and smart-city traffic management.

Generative AI deployment poses unprecedented challenges to content safety and privacy. However, existing defense mechanisms are often tailored to specific architectures (e.g., Diffusion Models or GANs), creating fragile "defense silos" that fail against heterogeneous generative threats. This paper identifies a fundamental optimization barrier in naive pixel-space ensemble strategies: due to divergent objective functions, pixel-level gradients from heterogeneous generators become statistically orthogonal, causing destructive interference. To overcome this, we observe that despite disparate low-level mechanisms, high-level feature representations of generated content exhibit alignment across architectures. Based on this, we propose the Architecture-Agnostic Targeted Feature Synergy (ATFS) framework. By introducing a target guidance image, ATFS reformulates multi-model defense as a unified feature space alignment task, enabling intrinsic gradient alignment without complex rectification. Extensive experiments show ATFS achieves SOTA protection in heterogeneous scenarios (e.g., Diffusion+GAN). It converges rapidly, reaching over 90% performance within 40 iterations, and maintains strong attack potency even under tight perturbation budgets. The framework seamlessly extends to unseen architectures (e.g., VQ-VAE) by switching the feature extractor, and demonstrates robust resistance to JPEG compression and scaling. Being computationally efficient and lightweight, ATFS offers a viable pathway to dismantle defense silos and enable universal generative security. Code and models are open-sourced for reproducibility.

Decompilation is foundational to binary analysis, yet conventional tools prioritize human readability over strict recompilability and verifiable runtime correctness. While recent LLM-based approaches attempt to refine decompiled pseudocode, they typically either optimize solely for readability or rely on static analysis for evaluation. This makes them prone to "semantic hallucinations" that compromise accuracy and fail to resolve actual runtime failures. For critical tasks like software modernization and vulnerability remediation, recovered code must not only compile but replicate the original binary's behavior. We present PCodeTrans, a feedback-driven framework that bridges the gap between decompilation, recompilation, and rigorous function-level dynamic validation. After extracting a minimal yet coherent context to guarantee recompilability, PCodeTrans employs an in situ substitutable engine to hot-swap the compiled function directly into the unmodified binary, natively preserving its authentic execution context and global dependencies. Guided by fine-grained differential tracing, PCodeTrans generates precise runtime feedback to iteratively guide an LLM in repairing semantic discrepancies. Evaluated on Coreutils and Binutils, PCodeTrans achieves unprecedented recovery performance when rectifying raw Hex-Rays outputs, attaining 100% function-level compilability on unstripped binaries alongside 99.55% and 99.89% test-validated behavioral consistency, respectively. In doing so, it resolves 76.56% and 79.74% of logic errors exposed by official test suites. Exhibiting exceptional resilience, PCodeTrans maintains over 96% behavioral consistency even on fully stripped binaries. By significantly outperforming all existing baselines, PCodeTrans paves a practical path to reliably translate decompiled pseudocode into compilable and executable equivalents.

We define a generic class of functions that captures most conceivable aggregations for Message-Passing Graph Neural Networks (MP-GNNs), and prove that any MP-GNN model with such aggregations induces only a polynomial number of equivalence classes on all graphs - while the number of non-isomorphic graphs is doubly-exponential (in number of vertices). Adding a familiar perspective, we observe that merely 2-iterations of Color Refinement (CR) induce at least an exponential number of equivalence classes, making the aforementioned MP-GNNs relatively infinitely weaker. Previous results state that MP-GNNs match full CR, however they concern a weak, 'non-uniform', notion of distinguishing-power where each graph size may required a different MP-GNN to distinguish graphs up to that size. Our results concern both distinguishing between non-equivariant vertices and distinguishing between non-isomorphic graphs.

Accurate day-ahead solar irradiance forecasting is essential for integrating solar energy into the power grid. However, it remains challenging due to the pronounced diurnal cycle and inherently complex cloud dynamics. Current methods either lack fine-scale resolution (e.g., numerical weather prediction, weather foundation models) or degrade at longer lead times (e.g., satellite extrapolation). We propose Baguan-solar, a two-stage multimodal framework that fuses forecasts from Baguan, a global weather foundation model, with high-resolution geostationary satellite imagery to produce 24- hour irradiance forecasts at kilometer scale. Its decoupled two-stage design first forecasts day-night continuous intermediates (e.g., cloud cover) and then infers irradiance, while its modality fusion jointly preserves fine-scale cloud structures from satellite and large-scale constraints from Baguan forecasts. Evaluated over East Asia using CLDAS as ground truth, Baguan-solar outperforms strong baselines (including ECMWF IFS, vanilla Baguan, and SolarSeer), reducing RMSE by 16.08% and better resolving cloud-induced transients. An operational deployment of Baguan-solar has supported solar power forecasting in an eastern province in China, since July 2025. Our code is accessible at https://github.com/DAMO-DI-ML/Baguansolar. git.

Toxicity detection mitigates the dissemination of toxic content (e.g., hateful comments, posts, and messages within online social actions) to safeguard a healthy online social environment. However, malicious users persistently develop evasive perturbations to disguise toxic content and evade detectors. Traditional detectors or methods are static over time and are inadequate in addressing these evolving evasion tactics. Thus, continual learning emerges as a logical approach to dynamically update detection ability against evolving perturbations. Nevertheless, disparities across perturbations hinder the detector's continual learning on perturbed text. More importantly, perturbation-induced noises distort semantics to degrade comprehension and also impair critical feature learning to render detection sensitive to perturbations. These amplify the challenge of continual learning against evolving perturbations. In this work, we present ContiGuard, the first framework tailored for continual learning of the detector on time-evolving perturbed text (termed continual toxicity detection) to enable the detector to continually update capability and maintain sustained resilience against evolving perturbations. Specifically, to boost the comprehension, we present an LLM-powered semantic enriching strategy, where we dynamically incorporate possible meaning and toxicity-related clues excavated by LLM into the perturbed text to improve the comprehension. To mitigate non-critical features and amplify critical ones, we propose a discriminability-driven feature learning strategy, where we strengthen discriminative features while suppressing the less-discriminative ones to shape a robust classification boundary for detection...

Road crashes remain a leading cause of preventable fatalities. Existing prediction models predominantly produce binary outcomes, which offer limited actionable insights for real-time driver feedback. These approaches often lack continuous risk quantification, interpretability, and explicit consideration of vulnerable road users (VRUs), such as pedestrians and cyclists. This research introduces SafeDriver-IQ, a framework that transforms binary crash classifiers into continuous 0-100 safety scores by combining national crash statistics with naturalistic driving data from autonomous vehicles. The framework fuses National Highway Traffic Safety Administration (NHTSA) crash records with Waymo Open Motion Dataset scenarios, engineers domain-informed features, and incorporates a calibration layer grounded in transportation safety literature. Evaluation across 15 complementary analyses indicates that the framework reliably differentiates high-risk from low-risk driving conditions with strong discriminative performance. Findings further reveal that 87% of crashes involve multiple co-occurring risk factors, with non-linear compounding effects that increase the risk to 4.5x baseline. SafeDriver-IQ delivers proactive, explainable safety intelligence relevant to advanced driver-assistance systems (ADAS), fleet management, and urban infrastructure planning. This framework shifts the focus from reactive crash counting to real-time risk prevention.

This paper studies the impact of retrieved ideological texts on the outputs of large language models (LLMs). While interest in understanding ideology in LLMs has recently increased, little attention has been given to this issue in the context of Retrieval-Augmented Generation (RAG). To fill this gap, we design an external knowledge source based on ideological loaded texts about COVID-19 treatments. Our corpus is based on 1,117 academic articles representing discourses about controversial and endorsed treatments for the disease. We propose a corpus linguistics framework, based on Lexical Multidimensional Analysis (LMDA), to identify the ideologies within the corpus. LLMs are tasked to answer questions derived from three identified ideological dimensions, and two types of contextual prompts are adopted: the first comprises the user question and ideological texts; and the second contains the question, ideological texts, and LMDA descriptions. Ideological alignment between reference ideological texts and LLMs' responses is assessed using cosine similarity for lexical and semantic representations. Results demonstrate that LLMs' responses based on ideological retrieved texts are more aligned with the ideology encountered in the external knowledge, with the enhanced prompt further influencing LLMs' outputs. Our findings highlight the importance of identifying ideological discourses within the RAG framework in order to mitigate not just unintended ideological bias, but also the risks of malicious manipulation of such models.

Multi-stream transformer architectures have recently been proposed as a promising direction for managing representation collapse and the vanishing gradient problem for residual connections, yet their internal mechanisms remain unexplored. In particular, the recently introduced Manifold-Constrained Hyper-Connections (mHC) architecture posits multiple residual streams with constrained interaction, but lacks in-depth mechanistic analysis. We present the first open-source mHC language model (https://huggingface.co/wgpeng/mhc-780m) and analyze the multiple-stream architecture with a suite of representation-level metrics and causal interventions to probe how parallel streams encode and utilize information. Specifically, we introduce a systematic stream ablation-and-rescue framework that enables direct causal comparison of residual streams during inference. Through targeted pairwise interventions and controlled recovery experiments, we distinguish functional redundancy from asymmetric utilization and reveal how information is distributed across streams beyond what is observable from representational similarity alone.

We propose a deep learning framework for COVID-19 detection and disease classification from chest CT scans that integrates both 2.5D and 3D representations to capture complementary slice-level and volumetric information. The 2.5D branch processes multi-view CT slices (axial, coronal, sagittal) using a DINOv3 vision transformer to extract robust visual features, while the 3D branch employs a ResNet-18 architecture to model volumetric context and is pretrained with Variance Risk Extrapolation (VREx) followed by supervised contrastive learning to improve cross-source robustness. Predictions from both branches are combined through logit-level ensemble inference. Experiments on the PHAROS-AIF-MIH benchmark demonstrate the effectiveness of the proposed approach: for binary COVID-19 detection, the ensemble achieves 94.48% accuracy and a 0.9426 Macro F1-score, outperforming both individual models, while for multi-class disease classification the 2.5D DINOv3 model achieves the best performance with 79.35% accuracy and a 0.7497 Macro F1-score. These results highlight the benefit of combining pretrained slice-based representations with volumetric modeling for robust multi-source medical imaging analysis. Code is available at https://github.com/HySonLab/PHAROS-AIF-MIH

We construct a cellular sheaf from any feedforward ReLU neural network by placing one vertex for each intermediate quantity in the forward pass and encoding each computational step - affine transformation, activation, output - as a restriction map on an edge. The restricted coboundary operator on the free coordinates is unitriangular, so its determinant is $1$ and the restricted Laplacian is positive definite for every activation pattern. It follows that the relative cohomology vanishes and the forward pass output is the unique harmonic extension of the boundary data. The sheaf heat equation converges exponentially to this output despite the state-dependent switching introduced by piecewise linear activations. Unlike the forward pass, the heat equation propagates information bidirectionally across layers, enabling pinned neurons that impose constraints in both directions, training through local discrepancy minimization without a backward pass, and per-edge diagnostics that decompose network behavior by layer and operation type. We validate the framework experimentally on small synthetic tasks, confirming the convergence theorems and demonstrating that sheaf-based training, while not yet competitive with stochastic gradient descent, obeys quantitative scaling laws predicted by the theory.

Dataset distillation, a training-aware data compression technique, has recently attracted increasing attention as an effective tool for mitigating costs of optimization and data storage. However, progress remains largely empirical. Mechanisms underlying the extraction of task-relevant information from the training process and the efficient encoding of such information into synthetic data points remain elusive. In this paper, we theoretically analyze practical algorithms of dataset distillation applied to the gradient-based training of two-layer neural networks with width $L$. By focusing on a non-linear task structure called multi-index model, we prove that the low-dimensional structure of the problem is efficiently encoded into the resulting distilled data. This dataset reproduces a model with high generalization ability for a required memory complexity of $\tildeΘ$$(r^2d+L)$, where $d$ and $r$ are the input and intrinsic dimensions of the task. To the best of our knowledge, this is one of the first theoretical works that include a specific task structure, leverage its intrinsic dimensionality to quantify the compression rate and study dataset distillation implemented solely via gradient-based algorithms.

Quantum computing provides the feasible multi-layered security challenges to classical blockchain systems. Whereas, quantum-secured blockchains relied on quantum key distribution (QKD) to establish secure channels can address this potential threat. This paper presents a scalable quantum-resistant blockchain architecture designed to address the connectivity and distance limitations of the QKD integrated quantum networks. By leveraging the twin-field (TF) QKD protocol within a measurement-device-independent (MDI) topology, the proposed framework can optimize the infrastructure complexity from quadratic to linear scaling. This architecture effectively integrates information-theoretic security with distributed consensus mechanisms, allowing the system to overcome the fundamental rate-loss limits inherent in traditional point-to-point links. The proposed scheme offers a theoretically sound and feasible solution for deploying large-scale and long-distance consortium.

Existing jailbreak defence frameworks for Large Vision-Language Models often suffer from a safety utility tradeoff, where strengthening safety inadvertently degrades performance on general visual-grounded reasoning tasks. In this work, we investigate whether safety and utility are inherently antagonistic objectives. We focus on a modality induced bias direction consistently observed across datasets, which arises from suboptimal coupling between the Large Language Model backbone and visual encoders. We further demonstrate that this direction undermines performance on both tasks. Leveraging this insight, we propose Two Birds, One Projection, an efficient inference time jailbreak defence that projects cross-modal features onto the null space of the identified bias direction to remove the corresponding components. Requiring only a single forward pass, our method effectively breaks the conventional tradeoff, simultaneously improving both safety and utility across diverse benchmarks.

Classical planning aims to find a sequence of actions, a plan, that maps a starting state into one of the goal states. If a trajectory appears to be leading to the goal, should we prioritise exploring it? Seminal work in goal recognition (GR) has defined GR in terms of a classical planning problem, adopting classical solvers and heuristics to recognise plans. We come full circle, and study the adoption and properties of GR-derived heuristics for seeking solutions to classical planning problems. We propose a new framework for assessing goal intention, which informs a new class of efficiently-computable heuristics. As a proof of concept, we derive two such heuristics, and show that they can already yield improvements for top-scoring classical planners. Our work provides foundational knowledge for understanding and deriving probabilistic intention-based heuristics for planning.

Deep Neural Networks demonstrate exceptional performance but remain vulnerable to adversarial perturbations, necessitating formal verification for safety-critical deployment. To address the computational complexity of this task, researchers often employ abstraction-refinement techniques that iteratively tighten an over-approximated model. While structural methods utilize Counterexample-Guided Abstraction Refine- ment, state-of-the-art dataflow verifiers typically rely on Branch-and-Bound to refine numerical convex relaxations. However, current dataflow approaches operate with blind refinement processes that rely on static heuristics and fail to leverage specific diagnostic information from verification failures. In this work, we argue that Branch-and-Bound should be reformulated as a Dataflow CEGAR loop where the spurious counterexample serves as a precise witness to local abstraction errors. We propose DRG-BaB, a framework that introduces the Directional Relaxation Gap heuristic to prioritize branching on neurons actively contributing to falsification in the abstract domain. By deriving a closed-form spurious counterexample directly from linear bounds, our method transforms generic search into targeted refinement. Experiments on high-dimensional benchmarks demonstrate that this approach significantly reduces search tree size and verification time compared to established baselines.

Transformer based diffusion and vision-language models have achieved remarkable success; yet, efficiently removing undesirable or sensitive information without retraining remains a central challenge for model safety and compliance. We introduce Ratio-Aware Zero/One-step Optimized Retentive unlearning (RAZOR), a lightweight, model-agnostic unlearning framework that generalizes forgetting updates to coordinated multi-layer and multi-head edits within transformer backbones. RAZOR identifies the most important layers and attention heads by measuring how much they contribute to forgetting the target data while preserving useful knowledge. Then, it updates these parts of the model using a carefully regularized rule to avoid harming overall performance. The set of edited components grows gradually, ensuring precise unlearning without over-editing or damaging unrelated capabilities. We evaluate RAZOR on CLIP, Stable Diffusion, and vision-language models (VLMs) using widely adopted unlearning benchmarks covering identity, style, and object erasure tasks. Our results show that RAZOR achieves highly accurate and stable forgetting, even under quantization. This approach offers stronger retention and better efficiency than prior methods. Notably, it also operates significant faster than conventional techniques. These results demonstrate that RAZOR is a practical and scalable solution for safe, adaptive unlearning in transformer-based vision models.

Deploying Deep Neural Networks (DNNs) on resource-constrained embedded systems requires aggressive model compression techniques like quantization and pruning. However, ensuring that the compressed model preserves the behavioral fidelity of the original design is a critical challenge in the safety-critical system design flow. Existing verification methods often lack scalability or fail to handle the architectural heterogeneity introduced by pruning. In this work, we propose SimCert, a probabilistic certification framework for verifying the behavioral similarity of compressed neural networks. Unlike worst-case analysis, SimCert provides quantitative safety guarantees with adjustable confidence levels. Our framework features: (1) A dual-network symbolic propagation method supporting both quantization and pruning; (2) A variance-aware bounding technique using Bernstein's inequality to tighten safety certificates; and (3) An automated verification toolchain. Experimental results on ACAS Xu and computer vision benchmarks demonstrate that SimCert outperforms state-of-the-art baselines.

Understanding cellular machinery requires atomic-scale reconstruction of large biomolecular assemblies. However, predicting the structures of these systems has been constrained by hardware memory requirements of models like AlphaFold 3, imposing a practical ceiling of a few thousand residues that can be processed on a single GPU. Here we present NVIDIA BioNeMo Fold-CP, a context parallelism framework that overcomes this barrier by distributing the inference and training pipelines of co-folding models across multiple GPUs. We use the Boltz models as open source reference architectures and implement custom multidimensional primitives that efficiently parallelize both the dense triangular updates and the irregular, data-dependent pattern of window-batched local attention. Our approach achieves efficient memory scaling; for an N-token input distributed across P GPUs, per-device memory scales as $O(N^2/P)$, enabling the structure prediction of assemblies exceeding 30,000 residues on 64 NVIDIA B300 GPUs. We demonstrate the scientific utility of this approach through successful developer use cases: Fold-CP enabled the scoring of over 90% of Comprehensive Resource of Mammalian protein complexes (CORUM) database, as well as folding of disease-relevant PI4KA lipid kinase complex bound to an intrinsically disordered region without cropping. By providing a scalable pathway for modeling massive systems with full global context, Fold-CP represents a significant step toward the realization of a virtual cell.

Enterprise software organizations accumulate critical institutional knowledge - architectural decisions, deployment procedures, compliance policies, incident playbooks - yet this knowledge remains trapped in formats designed for human interpretation. The bottleneck to effective agentic software development is not model capability but knowledge architecture. When any knowledge consumer - an autonomous AI agent, a newly onboarded engineer, or a senior developer - encounters an enterprise task without institutional context, the result is guesswork, correction cascades, and a disproportionate tax on senior engineers who must manually supply what others cannot infer. This paper introduces Knowledge Activation, a framework that specializes AI Skills - the open standard for agent-consumable knowledge - into structured, governance-aware Atomic Knowledge Units (AKUs) for institutional knowledge delivery. Rather than retrieving documents for interpretation, AKUs deliver action - ready specifications encoding what to do, which tools to use, what constraints to respect, and where to go next - so that agents act correctly and engineers receive institutionally grounded guidance without reconstructing organizational context from scratch. AKUs form a composable knowledge graph that agents traverse at runtime - compressing onboarding, reducing cross - team friction, and eliminating correction cascades. The paper formalizes the resource constraints that make this architecture necessary, specifies the AKU schema and deployment architecture, and grounds long - term maintenance in knowledge commons practice. Organizations that architect their institutional knowledge for the agentic era will outperform those that invest solely in model capability.

Target speech extraction (TSE) aims to recover a target speaker's voice from a mixture. While recent text-prompted approaches have shown promise, most approaches assume fully overlapped mixtures, limiting insight into behavior across realistic overlap ratios. We introduce VorTEX (Various overlap ratio for Target speech EXtraction), a text-prompted TSE architecture with a Decoupled Adaptive Multi-branch (DAM) Fusion block that separates primary extraction from auxiliary regularization pathways. To enable controlled analysis, we construct PORTE, a two-speaker dataset spanning overlap ratios from 0% to 100%. We further propose Suppression Ratio on Energy (SuRE), a diagnostic metric that detects suppression behavior not captured by conventional measures. Experiments show that existing models exhibit suppression or residual interference under overlap, whereas VorTEX achieves the highest separation fidelity across 20-100% overlap (e.g., 5.50 dB at 20% and 2.04 dB at 100%) while maintaining zero SuRE, indicating robust extraction without suppression-driven artifacts.

OpenReservoirComputing (ORC) is a Python library for reservoir computing (RC) written in JAX (Bradbury et al. 2018) and Equinox (Kidger and Garcia 2021). JAX is a Python library for high-performance numerical computing that enables automatic differentiation, just-in-time (JIT) compilation, and GPU/TPU acceleration, while Equinox is a neural network framework for JAX. RC is a form of machine learning that functions by lifting a low-dimensional sequence or signal into a high-dimensional dynamical system and training a simple, linear readout layer from the high-dimensional dynamics back to a lower-dimensional quantity of interest. The most common application of RC is time-series forecasting, where the goal is to predict a signal's future evolution. RC has achieved state-of-the-art performance on this task, particularly when applied to chaotic dynamical systems. In addition, RC approaches can be adapted to perform classification and control tasks. ORC provides both modular components for building custom RC models and built-in models for forecasting, classification, and control. By building on JAX and Equinox, ORC offers GPU acceleration, JIT compilation, and automatic vectorization. These capabilities make prototyping new models faster and enable larger and more powerful reservoir architectures. End-to-end differentiability also enables seamless integration with other deep learning models built with Equinox.

A critical failure mode of current lifelong agents is not lack of knowledge, but the inability to decide how to reason. When an agent encounters "Is this coin fair?" it must recognize whether to invoke frequentist hypothesis testing or Bayesian posterior inference - frameworks that are epistemologically incompatible. Mixing them produces not minor errors, but structural failures that propagate across decision chains. We formalize this as the universe routing problem: classifying questions into mutually exclusive belief spaces before invoking specialized solvers. Our key findings challenge conventional assumptions: (1) hard routing to heterogeneous solvers matches soft MoE accuracy while being 7x faster because epistemically incompatible frameworks cannot be meaningfully averaged; (2) a 465M-parameter router achieves a 2.3x smaller generalization gap than keyword-matching baselines, indicating semantic rather than surface-level reasoning; (3) when expanding to new belief spaces, rehearsal-based continual learning achieves zero forgetting, outperforming EWC by 75 percentage points, suggesting that modular epistemic architectures are fundamentally more amenable to lifelong learning than regularization-based approaches. These results point toward a broader architectural principle: reliable self-evolving agents may require an explicit epistemic control layer that governs reasoning framework selection.

We propose a machine-learning algorithm for Bayesian inverse problems in the function-space regime based on one-step generative transport. Building on the Mean Flows, we learn a fully conditional amortized sampler with a neural-operator backbone that maps a reference Gaussian noise to approximate posterior samples. We show that while white-noise references may be admissible at fixed discretization, they become incompatible with the function-space limit, leading to instability in inference for Bayesian problems arising from PDEs. To address this issue, we adopt a prior-aligned anisotropic Gaussian reference distribution and establish the Lipschitz regularity of the resulting transport. Our method is not distilled from MCMC: training relies only on prior samples and simulated partial and noisy observations. Once trained, it generates a $64\times64$ posterior sample in $\sim 10^{-3}$s, avoiding the repeated PDE solves of MCMC while matching key posterior summaries.

Accurate identification of protein-nucleotide binding sites is fundamental to deciphering molecular mechanisms and accelerating drug discovery. However, current computational methods often struggle with suboptimal performance due to inadequate feature representation and rigid fusion mechanisms, which hinder the effective exploitation of cross-task information synergy. To bridge this gap, we propose MTGA-MGE, a framework that integrates a Multi-Task Genetic Algorithm with Multi-Granularity Encoding to enhance binding site prediction. Specifically, we develop a Multi-Granularity Encoding (MGE) network that synergizes multi-scale convolutions and self-attention mechanisms to distill discriminative signals from high-dimensional, redundant biological data. To overcome the constraints of static fusion, a genetic algorithm is employed to adaptively evolve task-specific fusion strategies, thereby effectively improving model generalization. Furthermore, to catalyze collaborative learning, we introduce an External-Neighborhood Mechanism (ENM) that leverages biological similarities to facilitate targeted information exchange across tasks. Extensive evaluations on fifteen nucleotide datasets demonstrate that MTGA-MGE not only establishes a new state-of-the-art in data-abundant, high-resource scenarios but also maintains a robust competitive edge in rare, low-resource regimes, presenting a highly adaptive scheme for decoding complex protein-ligand interactions in the post-genomic era.

Modeling the reactive tempo of human conversation remains difficult because most audio-visual datasets portray isolated speakers delivering short monologues. We introduce \textbf{Face-to-Face with Jimmy Fallon (F2F-JF)}, a 70-hour, 14k-clip dataset of two-person talk-show exchanges that preserves the sequential dependency between a guest turn and the host's response. A semi-automatic pipeline combines multi-person tracking, speech diarization, and lightweight human verification to extract temporally aligned host/guest tracks with tight crops and metadata that are ready for downstream modeling. We showcase the dataset with a reactive, speech-driven digital avatar task in which the host video during $[t_1,t_2]$ is generated from their audio plus the guest's preceding video during $[t_0,t_1]$. Conditioning a MultiTalk-style diffusion model on this cross-person visual context yields small but consistent Emotion-FID and FVD gains while preserving lip-sync quality relative to an audio-only baseline. The dataset, preprocessing recipe, and baseline together provide an end-to-end blueprint for studying dyadic, sequential behavior, which we expand upon throughout the paper. Dataset and code will be made publicly available.

This paper proposes a novel hybrid model, termed GARCH-FIS, for recursive rolling multi-step forecasting of financial time series. It integrates a Fuzzy Inference System (FIS) with a Generalized Autoregressive Conditional Heteroskedasticity (GARCH) model to jointly address nonlinear dynamics and time-varying volatility. The core innovation is a dynamic parameter adaptation mechanism for the FIS, specifically activated within the multi-step forecasting cycle. In this process, the conditional volatility estimated by a rolling window GARCH model is continuously translated into a price volatility measure. At each forecasting step, this measure, alongside the updated mean of the sliding window data -- which now incorporates the most recent predicted price -- jointly determines the parameters of the FIS membership functions for the next prediction. Consequently, the granularity of the fuzzy inference adapts as the forecast horizon extends: membership functions are automatically widened during high-volatility market regimes to bolster robustness and narrowed during stable periods to enhance precision. This constitutes a fundamental advancement over a static one-step-ahead prediction setup. Furthermore, the model's fuzzy rule base is automatically constructed from data using the Wang-Mendel method, promoting interpretability and adaptability. Empirical evaluation, focused exclusively on multi-step forecasting performance across ten diverse financial assets, demonstrates that the proposed GARCH-FIS model significantly outperforms benchmark models -- including Support Vector Regression(SVR), Long Short-Term Memory networks(LSTM), and an ARIMA-GARCH econometric model -- in terms of predictive accuracy and stability, while effectively mitigating error accumulation in extended recursive forecasts.

Drug--target affinity prediction is pivotal for accelerating drug discovery, yet existing methods suffer from significant performance degradation in realistic cold-start scenarios (unseen drugs/targets/pairs), primarily driven by overfitting to training instances and information loss from irrelevant target sequences. In this paper, we propose LaPro-DTA, a framework designed to achieve robust and generalizable DTA prediction. To tackle overfitting, we devise a latent dual-view drug representation mechanism. It synergizes an instance-level view to capture fine-grained substructures with stochastic perturbation and a distribution-level view to distill generalized chemical scaffolds via semantic remapping, thereby enforcing the model to learn transferable structural rules rather than memorizing specific samples. To mitigate information loss, we introduce a salient protein feature extraction strategy using pattern-aware top-$k$ pooling, which effectively filters background noise and isolates high-response bioactive regions. Furthermore, a cross-view multi-head attention mechanism fuses these purified features to model comprehensive interactions. Extensive experiments on benchmark datasets demonstrate that LaPro-DTA significantly outperforms state-of-the-art methods, achieving an 8\% MSE reduction on the Davis dataset in the challenging unseen-drug setting, while offering interpretable insights into binding mechanisms.

Large language models (LLMs) have achieved remarkable performance across a wide range of tasks, but their inference efficiency remains a critical bottleneck due to rapidly growing parameters. Recent advances in dynamic computation allocation address this challenge by exploiting the highly uneven contributions of different tokens and layers, enabling selective execution that significantly reduces redundant computation while preserving model accuracy. However, existing hardware platforms and accelerators are primarily optimized for uniform, static execution, limiting their ability to efficiently support such dynamic inference patterns. In this work, we propose SkipOPU, an FPGA-based overlay processor that dynamically allocates computation across tokens and layers with high flexibility through a lightweight routing mechanism. First, we decouple reduction operations from element-wise computation in nonlinear modules and perform reductions incrementally, which enables both stages to be fused with adjacent linear operations (router or matrix multiplication) for effective latency hiding. Second, motivated by asymmetric sensitivity to numerical precision between activation and weight, we design a PE array that efficiently supports float-fixed hybrid execution. A novel DSP overpacking technique is introduced to maximize hardware utilization while minimizing resource overhead. Finally, we develop a proactive on-chip KV history buffer that exploits cross-layer KV invariance of pruned tokens, eliminating irregular HBM accesses during decoding and supplementing off-chip bandwidth through high-locality on-chip reuse. Experimental results demonstrate that SkipOPU on an AMD U280 FPGA outperforms GPU and other FPGA-based accelerators by 1.23x-3.83x in bandwidth efficiency for LLMs inference with dynamic computation allocation and can reduce up to 25.4% KV storage overhead across varying sequence lengths.

This paper presents Orthogonal Subspace Clustering (OSC), an innovative method for high-dimensional data clustering. We first establish a theoretical theorem proving that high-dimensional data can be decomposed into orthogonal subspaces in a statistical sense, whose form exactly matches the paradigm of Q-type factor analysis. This theorem lays a solid mathematical foundation for dimensionality reduction via matrix decomposition and factor analysis. Based on this theorem, we propose the OSC framework to address the "curse of dimensionality" -- a critical challenge that degrades clustering effectiveness due to sample sparsity and ineffective distance metrics. OSC integrates orthogonal subspace construction with classical clustering techniques, introducing a data-driven mechanism to select the subspace dimension based on cumulative variance contribution. This avoids manual selection biases while maximizing the retention of discriminative information. By projecting high-dimensional data into an uncorrelated, low-dimensional orthogonal subspace, OSC significantly improves clustering efficiency, robustness, and accuracy. Extensive experiments on various benchmark datasets demonstrate the effectiveness of OSC, with thorough analysis of evaluation metrics including Cluster Accuracy (ACC), Normalized Mutual Information (NMI), and Adjusted Rand Index (ARI) highlighting its advantages over existing methods.

Large Language Models (LLMs) are becoming a common way for humans to seek knowledge, yet their coverage and reliability vary widely. Especially for local language varieties, there are large asymmetries, e.g., information in local Wikipedia that is absent from the standard variant. However, little is known about how well LLMs perform under such information asymmetry, especially on closely related languages. We manually construct a novel challenge question-answering (QA) dataset that captures knowledge conveyed on a local Wikipedia page, which is absent from their higher-resource counterparts-covering Mandarin Chinese vs. Cantonese and German vs. Bavarian. Our experiments show that LLMs fail to answer questions about information only in local editions of Wikipedia. Providing context from lead sections substantially improves performance, with further gains possible via translation. Our topical, geographic annotations, and stratified evaluations reveal the usefulness of local Wikipedia editions as sources of both regional and global information. These findings raise critical questions about inclusivity and cultural coverage of LLMs.

Automatic Speech Recognition (ASR) performance is heavily dependent on the availability of large-scale, high-quality datasets. For low-resource languages, existing open-source ASR datasets often suffer from insufficient quality and inconsistent annotation, hindering the development of robust models. To address these challenges, we propose a novel and generalizable data aggregation and preprocessing pipeline designed to construct high-quality ASR datasets from diverse, potentially noisy, open-source sources. Our pipeline incorporates rigorous processing steps to ensure data diversity, balance, and the inclusion of crucial features like word-level timestamps. We demonstrate the effectiveness of our methodology by applying it to Vietnamese, resulting in a unified, high-quality 500-hour dataset that provides a foundation for training and evaluating state-of-the-art Vietnamese ASR systems. Our project page is available at https://github.com/qualcomm-ai-research/PhoASR.

Retrieval-Augmented Generation (RAG) enables large language models to use external knowledge, but outsourcing the RAG service raises privacy concerns for both data owners and users. Privacy-preserving RAG systems address these concerns by performing secure top-$k$ retrieval, which typically is secure sorting to identify relevant documents. However, existing systems face challenges supporting arbitrary $k$ due to their inability to change $k$, new security issues, or efficiency degradation with large $k$. This is a significant limitation because modern long-context models generally achieve higher accuracy with larger retrieval sets. We propose $p^2$RAG, a privacy-preserving RAG service that supports arbitrary top-$k$ retrieval. Unlike existing systems, $p^2$RAG avoids sorting candidate documents. Instead, it uses an interactive bisection method to determine the set of top-$k$ documents. For security, $p^2$RAG uses secret sharing on two semi-honest non-colluding servers to protect the data owner's database and the user's prompt. It enforces restrictions and verification to defend against malicious users and tightly bound the information leakage of the database. The experiments show that $p^2$RAG is 3--300$\times$ faster than the state-of-the-art PRAG for $k = 16$--$1024$.

Fine-tuning large models on edge devices is severely hindered by the memory-intensive backpropagation (BP) in standard frameworks like federated learning and split learning. While substituting BP with zeroth-order optimization can significantly reduce memory footprints, it typically suffers from prohibitively degraded convergence speed. To resolve this dilemma, we propose Hybrid-Order Split Federated Learning (HO-SFL). By reformulating the split learning process within a Lagrangian framework, HO-SFL decouples the optimization landscape: The server performs precise first-order updates (i.e., BP), whereas clients conduct memory-efficient zeroth-order optimization. This hybrid design not only eliminates the need for client-side BP but also enables dimension-free model aggregation, drastically lowering communication costs. Crucially, we provide a theoretical convergence analysis, demonstrating that HO-SFL mitigates the dimension-dependent convergence slowdown of zeroth-order optimization, achieving a convergence rate comparable to first-order methods. Extensive experiments on tasks across vision and language modalities validate that HO-SFL achieves convergence speeds comparable to first-order baselines while significantly reducing communication costs and client memory footprints.

Large Language Model (LLM)-based Collective Intelligence (CI) presents a promising approach to overcoming the data wall and continuously boosting the capabilities of LLM agents. However, there is currently no dedicated arena for evolving and benchmarking LLM-based CI. To address this gap, we introduce OpenHospital, an interactive arena where physician agents can evolve CI through interactions with patient agents. This arena employs a data-in-agent-self paradigm that rapidly enhances agent capabilities and provides robust evaluation metrics for benchmarking both medical proficiency and system efficiency. Experiments demonstrate the effectiveness of OpenHospital in both fostering and quantifying CI.

Optimizing complex systems, ranging from LLM prompts to multi-turn agents, traditionally requires labor-intensive manual iteration. We formalize this challenge as a stochastic generative optimization problem where a generative language model acts as the optimizer, guided by numerical rewards and text feedback to discover the best system. We introduce Prioritized Optimization with Local Contextual Aggregation (POLCA), a scalable framework designed to handle stochasticity in optimization -- such as noisy feedback, sampling minibatches, and stochastic system behaviors -- while effectively managing the unconstrained expansion of solution space. POLCA maintains a priority queue to manage the exploration-exploitation tradeoff, systematically tracking candidate solutions and their evaluation histories. To enhance efficiency, we integrate an $\varepsilon$-Net mechanism to maintain parameter diversity and an LLM Summarizer to perform meta-learning across historical trials. We theoretically prove that POLCA converges to near-optimal candidate solutions under stochasticity. We evaluate our framework on diverse benchmarks, including $τ$-bench, HotpotQA (agent optimization), VeriBench (code translation) and KernelBench (CUDA kernel generation). Experimental results demonstrate that POLCA achieves robust, sample and time-efficient performance, consistently outperforming state-of-the-art algorithms in both deterministic and stochastic problems. The codebase for this work is publicly available at https://github.com/rlx-lab/POLCA.

For classification tasks, the performance of a deep neural network is determined by the structure of its decision boundary, whose geometry directly affects essential properties of the model, including accuracy and robustness. Motivated by a classical tube formula due to Weyl, we introduce a method to measure the decision boundary of a neural network through local surface volumes, providing a theoretically justifiable and efficient measure enabling a geometric interpretation of the effectiveness of the model applicable to the high dimensional feature spaces considered in deep learning. A smaller surface volume is expected to correspond to lower model complexity and better generalisation. We verify, on a number of image processing tasks with convolutional architectures that decision boundary volume is inversely proportional to classification accuracy. Meanwhile, the relationship between local surface volume and generalisation for fully connected architecture is observed to be less stable between tasks. Therefore, for network architectures suited to a particular data structure, we demonstrate that smoother decision boundaries lead to better performance, as our intuition would suggest.

Logical Access (LA) attacks, also known as audio deepfake attacks, use Text-to-Speech (TTS) or Voice Conversion (VC) methods to generate spoofed speech data. This can represent a serious threat to Automatic Speaker Verification (ASV) systems, as intruders can use such attacks to bypass voice biometric security. In this study, we investigate the correlation between speech quality and the performance of audio spoofing detection systems (i.e., LA task). For that, the performance of two enhancement algorithms is evaluated based on two perceptual speech quality measures, namely Perceptual Evaluation of Speech Quality (PESQ) and Speech-to-Reverberation Modulation Ratio (SRMR), and in respect to their impact on the audio spoofing detection system. We adopted the LA dataset, provided in the ASVspoof 2019 Challenge, and corrupted its test set with different Signal-to-Noise Ratio (SNR) levels, while leaving the training data untouched. Enhancement was applied to attenuate the detrimental effects of noisy speech, and the performances of two models, Speech Enhancement Generative Adversarial Network (SEGAN) and Metric-Optimized Generative Adversarial Network Plus (MetricGAN+), were compared. Although we expect that speech quality will correlate well with speech applications' performance, it can also have as a side effect on downstream tasks if unwanted artifacts are introduced or relevant information is removed from the speech signal. Our results corroborate with this hypothesis, as we found that the enhancement algorithm leading to the highest speech quality scores, MetricGAN+, provided the lowest Equal Error Rate (EER) on the audio spoofing detection task, whereas the enhancement method with the lowest speech quality scores, SEGAN, led to the lowest EER, thus leading to better performance on the LA task.

Geometric data augmentation is widely used in segmentation pipelines and typically assumes that polygon annotations represent simply connected regions. However, in structured domains such as architectural floorplan analysis, ring-type regions are often encoded as a single cyclic polygon chain connecting outer and inner boundaries. During augmentation, clipping operations may remove intermediate vertices and disrupt this cyclic connectivity, breaking the structural relationship between the boundaries. In this work, we introduce an order-preserving polygon augmentation strategy that performs transformations in mask space and then projects surviving vertices back into index-space to restore adjacency relations. This repair maintains the original traversal order of the polygon and preserves topological consistency with minimal computational overhead. Experiments demonstrate that the approach reliably restores connectivity, achieving near-perfect Cyclic Adjacency Preservation (CAP) across both single and compound augmentations.

We study supervisory switching control for partially-observed linear dynamical systems. The objective is to identify and deploy the best controller for the unknown system by periodically selecting among a collection of $N$ candidate controllers, some of which may destabilize the underlying system. While classical estimator-based supervisory control guarantees asymptotic stability, it lacks quantitative finite-time performance bounds. Conversely, current non-asymptotic methods in both online learning and system identification require restrictive assumptions that are incompatible in a control setting, such as system stability, which preclude testing potentially unstable controllers. To bridge this gap, we propose a novel, non-asymptotic analysis of supervisory control that adapts multi-armed bandit algorithms to address these control-theoretic challenges. Our data-driven algorithm evaluates candidate controllers via scoring criteria that leverage system observability to isolate the effects of historical states, enabling both detection of destabilizing controllers and accurate system identification. We present two algorithmic variants with dimension-free, finite-time guarantees, where each identifies the most suitable controller in $\mathcal{O}(N \log N)$ steps, while simultaneously achieving finite $L_2$-gain with respect to system disturbances.

Large language models (LLMs) achieve impressive scores on standard benchmarks yet routinely fail questions that any human would answer correctly in seconds. We introduce BrainBench, a benchmark of 100 brainteaser questions spanning 20 carefully designed categories, each targeting a specific commonsense reasoning failure mode in LLMs. Categories range from implicit physical constraints ("Should I walk or drive my rental car to the return lot?") to semantic scope tricks and default assumption hijacks. We evaluate eight frontier models -- four from the Claude family and four from the GPT family -- using a zero-shot protocol with 10 independent runs per question. The best model, Claude Opus 4.6 with extended thinking, achieves only 80.3% accuracy; the worst, GPT-4o, scores 39.7%. Even top-performing models exhibit a 6-16 percentage-point gap between accuracy and consistency, revealing stochastic reasoning. Cross-lingual evaluation in Chinese shows most models degrade by 2-8 percentage points, confirming that these failures reflect reasoning deficits rather than language-specific artifacts. BrainBench provides a fine-grained diagnostic tool for identifying where and why LLMs substitute surface heuristics for genuine commonsense reasoning.

Current online translation services require sending user text to cloud servers, posing a risk of privacy leakage when the text contains sensitive information. This risk hinders the application of online translation services in privacy-sensitive scenarios. One way to mitigate this risk for online translation services is introducing privacy protection mechanisms targeting the inference stage of translation models. However, compared to subfields of NLP like text classification and summarization, the machine translation research community has limited exploration of privacy protection during the inference stage. There is no clearly defined privacy protection task for the inference stage, dedicated evaluation datasets and metrics, and reference benchmark methods. The absence of these elements has seriously constrained researchers' in-depth exploration of this direction. To bridge this gap, this paper proposes a novel "Privacy-Preserving Machine Translation" (PPMT) task, aiming to protect the private information in text during the model inference stage. For this task, we constructed three benchmark test datasets, designed corresponding evaluation metrics, and proposed a series of benchmark methods as a starting point for this task. The definition of privacy is complex and diverse. Considering that named entities often contain a large amount of personal privacy and commercial secrets, we have focused our research on protecting only the named entity's privacy in the text. We expect this research work will provide a new perspective and a solid foundation for the privacy protection problem in machine translation.

Headedness is widely used as an organizing device in syntactic analysis, yet constituency treebanks rarely encode it explicitly and most processing pipelines recover it procedurally via percolation rules. We treat this notion of constituent headedness as an explicit representational layer and learn it as a supervised prediction task over aligned constituency and dependency annotations, inducing supervision by defining each constituent head as the dependency span head. On aligned English and Chinese data, the resulting models achieve near-ceiling intrinsic accuracy and substantially outperform Collins-style rule-based percolation. Predicted heads yield comparable parsing accuracy under head-driven binarization, consistent with the induced binary training targets being largely equivalent across head choices, while increasing the fidelity of deterministic constituency-to-dependency conversion and transferring across resources and languages under simple label-mapping interfaces.

Recent advances in Multimodal Large Language Models (MLLMs) have shown impressive reasoning capabilities across vision-language tasks, yet still face the challenge of compute-difficulty mismatch. Through empirical analyses, we identify that existing decoding methods may waste compute on easy cases while underserving hard ones, affecting both model effectiveness and efficiency. To address this issue, we first develop a theoretical framework that links sampling coverage, instance difficulty, and residual risk. Our analysis reveals that multimodal reasoning exhibits a heavy-tailed difficulty distribution; a small subset of hard or ambiguous samples dominates the residual failure probability. Based on this insight, we propose Coverage-Aware Multimodal Decoding (CAMD), an adaptive inference mechanism that dynamically allocates computation according to estimated uncertainty. CAMD integrates evidence-weighted scoring, posterior coverage estimation, and sequential Bayesian updating to balance efficiency and reliability under a limited token budget. Experiments on various benchmark datasets and baselines demonstrate the effectiveness and advantages of our approach.

Future trajectory prediction of a tracked pedestrian from an egocentric perspective is a key task in areas such as autonomous driving and robot navigation. The challenge of this task lies in the complex dynamic relative motion between the ego-camera and the tracked pedestrian. To address this challenge, we propose an ego-motion-guided trajectory prediction network based on the Mamba model. Firstly, two Mamba models are used as encoders to extract pedestrian motion and ego-motion features from pedestrian movement and ego-vehicle movement, respectively. Then, an ego-motion guided Mamba decoder that explicitly models the relative motion between the pedestrian and the vehicle by integrating pedestrian motion features as historical context with ego-motion features as guiding cues to capture decoded features. Finally, the future trajectory is generated from the decoded features corresponding to the future timestamps. Extensive experiments demonstrate the effectiveness of the proposed model, which achieves state-of-the-art performance on the PIE and JAAD datasets.

Learning solution operators of partial differential equations (PDEs) from data has emerged as a promising route to fast surrogate models in multi-query scientific workflows. However, for geometric PDEs whose inputs and outputs transform under changes of local frame (gauge), many existing operator-learning architectures remain representation-dependent, brittle under metric perturbations, and sensitive to discretization changes. We propose Gauge-Equivariant Intrinsic Neural Operators (GINO), a class of neural operators that parameterize elliptic solution maps primarily through intrinsic spectral multipliers acting on geometry-dependent spectra, coupled with gauge-equivariant nonlinearities. This design decouples geometry from learnable functional dependence and enforces consistency under frame transformations. We validate GINO on controlled problems on the flat torus ($\mathbb{T}^2$), where ground-truth resolvent operators and regularized Helmholtz--Hodge decompositions admit closed-form Fourier representations, enabling theory-aligned diagnostics. Across experiments E1--E6, GINO achieves low operator-approximation error, near machine-precision gauge equivariance, robustness to structured metric perturbations, strong cross-resolution generalization with small commutation error under restriction/prolongation, and structure-preserving performance on a regularized exact/coexact decomposition task. Ablations further link the smoothness of the learned spectral multiplier to stability under geometric perturbations. These results suggest that enforcing intrinsic structure and gauge equivariance yields operator surrogates that are more geometry-consistent and discretization-robust for elliptic PDEs on form-valued fields.

As large language models (LLMs) are increasingly considered for automated assessment and feedback, understanding when LLM marking can be trusted is essential. We evaluate LLM-as-a-judge marking across three physics assessment formats - structured questions, written essays, and scientific plots - comparing GPT-5.2, Grok 4.1, Claude Opus 4.5, DeepSeek-V3.2, Gemini Pro 3, and committee aggregations against human markers under blind, solution-provided, false-solution, and exemplar-anchored conditions. For $n=771$ blind university exam questions, models achieve fractional mean absolute errors (fMAE) $\approx 0.22$ with robust discriminative validity (Spearman $ρ> 0.6$). For secondary and university structured questions ($n=1151$), providing official solutions reduces MAE and strengthens validity (committee $ρ= 0.88$); false solutions degrade absolute accuracy but leave rank ordering largely intact (committee $ρ= 0.77$; individual models $ρ\geq 0.59$). Essay marking behaves fundamentally differently. Across $n=55$ scripts ($n=275$ essays), blind AI marking is harsher and more variable than human marking, with discriminative validity already poor ($ρ\approx 0.1$). Adding a mark scheme does not improve discrimination ($ρ\approx 0$; all confidence intervals include zero). Anchored exemplars shift the AI mean close to the human mean and compress variance below the human standard deviation, but discriminative validity remains near-zero - distributional agreement can occur without valid discrimination. For code-based plot elements ($n=1400$), models achieve exceptionally high discriminative validity ($ρ> 0.84$) with near-linear calibration. Across all task types, validity tracks criterion-referenceability - the extent to which a task maps to explicit, observable grading features - and benchmark reliability, rather than raw model capability.

Large language models (LLMs) facilitate the development of autonomous agents. As a core component of such agents, task planning aims to decompose complex natural language requests into concrete, solvable sub-tasks. Since LLM-generated plans are frequently prone to hallucinations and sensitive to long-context prom-pts, recent research has introduced plan verifiers to identify and correct potential flaws. However, most existing approaches still rely on an LLM as the verifier via additional prompting for plan review or self-reflection. LLM-based verifiers can be misled by plausible narration and struggle to detect failures caused by structural relations across steps, such as type mismatches, missing intermediates, or broken dependencies. To address these limitations, we propose a graph-based verifier for LLM task planning. Specifically, the proposed method has four major components: Firstly, we represent a plan as a directed graph with enriched attributes, where nodes denote sub-tasks and edges encode execution order and dependency constraints. Secondly, a graph neural network (GNN) then performs structural evaluation and diagnosis, producing a graph-level plausibility score for plan acceptance as well as node/edge-level risk scores to localize erroneous regions. Thirdly, we construct controllable perturbations from ground truth plan graphs, and automatically generate training data with fine-grained annotations. Finally, guided by the feedback from our GNN verifier, we enable an LLM to conduct local edits (e.g., tool replacement or insertion) to correct the plan when the graph-level score is insufficient. Extensive experiments across diverse datasets, backbone LLMs, and planners demonstrate that our GNNVerifier achieves significant gains in improving plan quality. Our data and code is available at https://github.com/BUPT-GAMMA/GNNVerifier.

Next-generation IoT applications increasingly span across autonomous administrative entities, necessitating silo-cooperative scheduling to leverage diverse computational resources while preserving data privacy. However, realizing efficient cooperation faces significant challenges arising from infrastructure heterogeneity, Non-IID workload shifts, and the inherent risks of adversarial environments. Existing approaches, relying predominantly on centralized coordination or independent learning, fail to address the incompatibility of state-action spaces across heterogeneous silos and lack robustness against malicious attacks. This paper proposes DeFRiS, a Decentralized Federated Reinforcement Learning framework for robust and scalable Silo-cooperative IoT application scheduling. DeFRiS integrates three synergistic innovations: (i) an action-space-agnostic policy utilizing candidate resource scoring to enable seamless knowledge transfer across heterogeneous silos; (ii) a silo-optimized local learning mechanism combining Generalized Advantage Estimation (GAE) with clipped policy updates to resolve sparse delayed reward challenges; and (iii) a Dual-Track Non-IID robust decentralized aggregation protocol leveraging gradient fingerprints for similarity-aware knowledge transfer and anomaly detection, and gradient tracking for optimization momentum. Extensive experiments on a distributed testbed with 20 heterogeneous silos and realistic IoT workloads demonstrate that DeFRiS significantly outperforms state-of-the-art baselines, reducing average response time by 6.4% and energy consumption by 7.2%, while lowering tail latency risk (CVaR$_{0.95}$) by 10.4% and achieving near-zero deadline violations. Furthermore, DeFRiS achieves over 3 times better performance retention as the system scales and over 8 times better stability in adversarial environments compared to the best-performing baseline.

Game UI design requires consistent visual assets across rarity tiers yet remains a predominantly manual process. We present GameUIAgent, an LLM-powered agentic framework that translates natural language descriptions into editable Figma designs via a Design Spec JSON intermediate representation. A six-stage neuro-symbolic pipeline combines LLM generation, deterministic post-processing, and a Vision-Language Model (VLM)-guided Reflection Controller (RC) for iterative self-correction with guaranteed non-regressive quality. Evaluated across 110 test cases, three LLMs, and three UI templates, cross-model analysis establishes a game-domain failure taxonomy (rarity-dependent degradation; visual emptiness) and uncovers two key empirical findings. A Quality Ceiling Effect (Pearson r=-0.96, p<0.01) suggests that RC improvement is bounded by headroom below a quality threshold -- a visual-domain counterpart to test-time compute scaling laws. A Rendering-Evaluation Fidelity Principle reveals that partial rendering enhancements paradoxically degrade VLM evaluation by amplifying structural defects. Together, these results establish foundational principles for LLM-driven visual generation agents in game production.

How safety supervision is written may matter more than the explicit identity content it contains. We study low-data LoRA safety fine-tuning with four supervision formats built from the same core safety rules: constitutional rules (A), creed-style identity framing (B), a B-matched creed condition with a worldview/confession identity-maintenance tail (C), and a matched non-identity condition (D). Across three instruction-tuned model families (Llama 3.1 8B, Qwen2.5 7B, and Gemma 3 4B), we evaluate HarmBench using a reconciled dual-judge pipeline combining Bedrock-hosted DeepSeek v3.2 and Sonnet 4.6, with disagreement and boundary cases manually resolved. The non-identity condition D is the strongest group on all three model families on the full 320-behavior HarmBench set, reaching 74.4% refusal on Llama, 76.9% on Gemma, and 74.1% on Qwen. By comparison, creed-style framing (B) improves over plain constitutional rules (A) on Llama and Gemma, but remains substantially below D, yielding an overall descriptive ordering of $D > B > C \geq A > baseline$. This provides a bounded empirical challenge to a strong version of the identity-framing hypothesis: explicit creed-style identity language is not necessary for the strongest gains observed here. Capability evaluations on MMLU and ARC-Challenge show no meaningful trade-off across conditions.

Early identification of patients at risk for clinical deterioration in the intensive care unit (ICU) remains a critical challenge. Delayed recognition of impending adverse events, including mortality, vasopressor initiation, and mechanical ventilation, contributes to preventable morbidity and mortality. We present a multimodal deep learning approach that combines structured time-series data (vital signs and laboratory values) with unstructured clinical notes to predict patient deterioration within 24 hours. Using the MIMIC-IV database, we constructed a cohort of 74,822 ICU stays and generated 5.7 million hourly prediction samples. Our architecture employs a bidirectional LSTM encoder for temporal patterns in physiologic data and ClinicalBERT embeddings for clinical notes, fused through a cross-modal attention mechanism. We also present a systematic review of existing approaches to ICU deterioration prediction, identifying 31 studies published between 2015 and 2024. Most existing models rely solely on structured data and achieve area under the curve (AUC) values between 0.70 and 0.85. Studies incorporating clinical notes remain rare but show promise for capturing information not present in structured fields. Our multimodal model achieves a test AUROC of 0.7857 and AUPRC of 0.1908 on 823,641 held-out samples, with a validation-to-test gap of only 0.6 percentage points. Ablation analysis validates the multimodal approach: clinical notes improve AUROC by 2.5 percentage points and AUPRC by 39.2% relative to a structured-only baseline, while deep learning models consistently outperform classical baselines (XGBoost AUROC: 0.7486, logistic regression: 0.7171). This work contributes both a thorough review of the field and a reproducible multimodal framework for clinical deterioration prediction.

Most protein families have fewer than 100 known members, a regime where deep generative models overfit or collapse. We propose stochastic attention (SA), a training-free sampler that treats the modern Hopfield energy over a protein alignment as a Boltzmann distribution and draws samples via Langevin dynamics. The score function is a closed-form softmax attention operation requiring no training, no pretraining data, and no GPU, with cost linear in alignment size. Across eight Pfam families, SA generates sequences with low amino acid compositional divergence, substantial novelty, and structural plausibility confirmed by ESMFold and AlphaFold2. Generated sequences fold more faithfully to canonical family structures than natural members in six of eight families. Against profile HMMs, EvoDiff, and the MSA Transformer, which produce sequences that drift far outside the family, SA maintains 51 to 66 percent identity while remaining novel, in seconds on a laptop. The critical temperature governing generation is predicted from PCA dimensionality alone, enabling fully automatic operation. Controls confirm SA encodes correlated substitution patterns, not just per-position amino acid frequencies.

Large language models (LLMs) have demonstrated remarkable performance across a wide range of tasks and domains, with data playing a central role in enabling these advances. Despite this success, the preparation and effective utilization of the massive datasets required for LLM training remain major bottlenecks. In current practice, LLM training data is often constructed using ad hoc scripts, and there is still a lack of mature, agent-based data preparation systems that can automatically construct robust and reusable data workflows, thereby freeing data scientists from repetitive and error-prone engineering efforts. Moreover, once collected, datasets are often consumed largely in their entirety during training, without systematic mechanisms for data selection, mixture optimization, or reweighting. To address these limitations, we advocate two complementary research directions. First, we propose building a robust, agent-based automatic data preparation system that supports automated workflow construction and scalable data management. Second, we argue for a unified data-model interaction training system in which data is dynamically selected, mixed, and reweighted throughout the training process, enabling more efficient, adaptive, and performance-aware data utilization. Finally, we discuss the remaining challenges and outline promising directions for future research and system development.

Recent advances in time series foundation models (TSFMs) demonstrate strong expressive capacity through large-scale pretraining across diverse time series domains. Zero-shot time series forecasting with TSFMs, however, exhibits limited generalization to unseen datasets, which retrieval-augmented forecasting addresses by leveraging an external knowledge base. Existing approaches rely on a fixed number of retrieved samples that may introduce irrelevant information. To this end, we propose Cross-RAG, a zero-shot retrieval-augmented forecasting framework that selectively attends to query-relevant retrieved samples. Cross-RAG models input-level relevance between the query and retrieved samples via query-retrieval cross-attention, while jointly incorporating information from the query and retrieved samples. Extensive experiments demonstrate that Cross-RAG consistently improves zero-shot forecasting performance across various TSFMs and RAG methods, and additional analyses confirm its effectiveness across diverse retrieval scenarios. Code is available at https://github.com/seunghan96/cross-rag/.

Computer-using agents (CUAs) act directly on graphical user interfaces, yet their perception of the screen is often unreliable. Existing work largely treats these failures as performance limitations, asking whether an action succeeds, rather than whether the agent is acting on the correct object at all. We argue that this is fundamentally a security problem. We formalize the visual confused deputy: a failure mode in which an agent authorizes an action based on a misperceived screen state, due to grounding errors, adversarial screenshot manipulation, or time-of-check-to-time-of-use (TOCTOU) races. This gap is practically exploitable: even simple screen-level manipulations can redirect routine clicks into privileged actions while remaining indistinguishable from ordinary agent mistakes. To mitigate this threat, we propose the first guardrail that operates outside the agent's perceptual loop. Our method, dual-channel contrastive classification, independently evaluates (1) the visual click target and (2) the agent's reasoning about the action against deployment-specific knowledge bases, and blocks execution if either channel indicates risk. The key insight is that these two channels capture complementary failure modes: visual evidence detects target-level mismatches, while textual reasoning reveals dangerous intent behind visually innocuous controls. Across controlled attacks, real GUI screenshots, and agent traces, the combined guardrail consistently outperforms either channel alone. Our results suggest that CUA safety requires not only better action generation, but independent verification of what the agent believes it is clicking and why. Materials are provided\footnote{Model, benchmark, and code: https://github.com/vllm-project/semantic-router}.

Frozen-backbone transfer with Vision Transformers faces two under-addressed issues: optimization instability when adapters are naively inserted into a fixed feature extractor, and the absence of principled guidance for setting adapter capacity. We introduce AdapterTune, which augments each transformer block with a residual low-rank bottleneck whose up-projection is zero-initialized, guaranteeing that the adapted network starts exactly at the pretrained function and eliminates early-epoch representation drift. On the analytical side, we formalize adapter rank as a capacity budget for approximating downstream task shifts in feature space. The resulting excess-risk decomposition predicts monotonic but diminishing accuracy gains with increasing rank, an ``elbow'' behavior we confirm through controlled sweeps. We evaluate on 9 datasets and 3 backbone scales with multi-seed reporting throughout. On a core 5 dataset transfer suite, AdapterTune improves top-1 accuracy over head-only transfer by +14.9 points on average while training only 0.92 of the parameters required by full fine-tuning, and outperforms full fine-tuning on 10 of 15 dataset-backbone pairs. Across the full benchmark, AdapterTune improves over head-only transfer on every dataset-backbone pair tested. Ablations on rank, placement, and initialization isolate each design choice. The code is available at: https://github.com/salimkhazem/adaptertune

Diffusion models operate in a reflexive System 1 mode, constrained by a fixed, content-agnostic sampling schedule. This rigidity arises from the curse of state dimensionality, where the combinatorial explosion of possible states in the high-dimensional noise manifold renders explicit trajectory planning intractable and leads to systematic computational misallocation. To address this, we introduce Chain-of-Trajectories (CoTj), a train-free framework enabling System 2 deliberative planning. Central to CoTj is Diffusion DNA, a low-dimensional signature that quantifies per-stage denoising difficulty and serves as a proxy for the high-dimensional state space, allowing us to reformulate sampling as graph planning on a directed acyclic graph. Through a Predict-Plan-Execute paradigm, CoTj dynamically allocates computational effort to the most challenging generative phases. Experiments across multiple generative models demonstrate that CoTj discovers context-aware trajectories, improving output quality and stability while reducing redundant computation. This work establishes a new foundation for resource-aware, planning-based diffusion modeling. The code is available at https://github.com/UnicomAI/CoTj.

Large language models and AI agents have recently shown promise in automating software performance optimization, but existing approaches predominantly rely on local, syntax-driven code transformations. This limits their ability to reason about program behavior and capture whole system performance interactions. As modern software increasingly comprises interacting components - such as microservices, databases, and shared infrastructure - effective code optimization requires reasoning about program structure and system architecture beyond individual functions or files. This paper explores the feasibility of whole system optimization for microservices. We introduce a multi-agent framework that integrates control-flow and data-flow representations with architectural and cross-component dependency signals to support system-level performance reasoning. The proposed system is decomposed into coordinated agent roles - summarization, analysis, optimization, and verification - that collaboratively identify cross-cutting bottlenecks and construct multi-step optimization strategies spanning the software stack. We present a proof-of-concept on a microservice-based system that illustrates the effectiveness of our proposed framework, achieving a 36.58% improvement in throughput and a 27.81% reduction in average response time.

Machine learning interatomic potentials (MLIPs) have become widely used tools in atomistic simulations. For much of the history of this field, the most commonly employed architectures were based on short-ranged atomic energy contributions, and the assumption of locality still persists in many modern foundation models. While this approach has enabled efficient and accurate modelling for many use cases, it poses intrinsic limitations for systems where long-range electrostatics, charge transfer, or induced polarization play a central role. A growing body of work has proposed extensions that incorporate electrostatic effects, ranging from locally predicted atomic charges to self-consistent models. While these models have demonstrated success for specific examples, their underlying assumptions, and fundamental limitations are not yet well understood. In this work, we present a framework for treating electrostatics in MLIPs by viewing existing models as coarse-grained approximations to density functional theory (DFT). This perspective makes explicit the approximations involved, clarifies the physical meaning of the learned quantities, and reveals connections and equivalences between several previously proposed models. Using this formalism, we identify key design choices that define a broader design space of self-consistent electrostatic MLIPs. We implement salient points in this space using the MACE architecture and a shared representation of the charge density, enabling controlled comparisons between different approaches. Finally, we evaluate these models on two instructive test cases: metal-water interfaces, which probe the contrasting electrostatic response of conducting and insulating systems, and charged vacancies in silicon dioxide. Our results highlight the limitations of existing approaches and demonstrate how more expressive self-consistent models are needed to resolve failures.

Cooperative multi-robot missions often require teams of robots to traverse environments where traversal risk evolves due to adversary patrols or shifting hazards with stochastic dynamics. While support coordination - where robots assist teammates in traversing risky regions - can significantly reduce mission costs, its effectiveness depends on the team's ability to anticipate future risk. Existing support-based frameworks assume static risk landscapes and therefore fail to account for predictable temporal trends in risk evolution. We propose a forecast-aware cooperative planning framework that integrates stochastic risk forecasting with anticipatory support allocation on temporal graphs. By modeling adversary dynamics as a first-order Markov stay-move process over graph edges, we propagate the resulting edge-occupancy probabilities forward in time to generate time-indexed edge-risk forecasts. These forecasts guide the proactive allocation of support positions to forecasted risky edges for effective support coordination, while also informing joint robot path planning. Experimental results demonstrate that our approach consistently reduces total expected team cost compared to non-anticipatory baselines, approaching the performance of an oracle planner.

Predicting how materials behave under realistic conditions requires understanding the statistical distribution of atomic configurations on crystal lattices, a problem central to alloy design, catalysis, and the study of phase transitions. Traditional Markov-chain Monte Carlo sampling suffers from slow convergence and critical slowing down near phase transitions, motivating the use of generative models that directly learn the thermodynamic distribution. Existing autoregressive models (ARMs), however, generate configurations in a fixed sequential order and incur high memory and training costs, limiting their applicability to realistic systems. Here, we develop a framework combining any-order ARMs, which generate configurations flexibly by conditioning on any known subset of lattice sites, with marginalization models (MAMs), which approximate the probability of any partial configuration in a single forward pass and substantially reduce memory requirements. This combination enables models trained on smaller lattices to be reused for sampling larger systems, while supporting expressive Transformer architectures with lattice-aware positional encodings at manageable computational cost. We demonstrate that Transformer-based any-order MAMs achieve more accurate free energies than multilayer perceptron-based ARMs on both the two-dimensional Ising model and CuAu alloys, faithfully capturing phase transitions and critical behavior. Overall, our framework scales from $10 \times 10$ to $20 \times 20$ Ising systems and from $2 \times 2 \times 4$ to $4 \times 4 \times 8$ CuAu supercells at reduced computational cost compared to conventional sampling methods.

Rapid structural damage assessment from remote sensing imagery is essential for timely disaster response. Within human-machine systems (HMS) for disaster management, automated damage detection provides decision-makers with actionable situational awareness. However, models trained on multi-disaster benchmarks often underperform in unseen geographic regions due to domain shift - a distributional mismatch between training and deployment data that undermines human trust in automated assessments. We explore a two-stage ensemble approach using supervised domain adaptation (SDA) for building damage classification across four severity classes. The pipeline adapts the xView2 first-place method to the Ida-BD dataset using SDA and systematically investigates the effect of individual augmentation components on classification performance. Comprehensive ablation experiments on the unseen Ida-BD test split demonstrate that SDA is indispensable: removing it causes damage detection to fail entirely. Our pipeline achieves the most robust performance using SDA with unsharp-enhanced RGB input, attaining a Macro-F1 of 0.5552. These results underscore the critical role of domain adaptation in building trustworthy automated damage assessment modules for HMS-integrated disaster response.

The relationship between intuitionistic or intermediate logics and logic programming has been extensively studied, prominently featuring Pearce's equilibrium logic and Osorio's safe beliefs. Equilibrium logic admits a fixpoint characterization based on the logic of here-and-there, akin to theory completion in default and autoepistemic logics. Safe beliefs are similarly defined via a fixpoint operator, albeit under the semantics of intuitionistic or other intermediate logics. In this paper, we investigate the logical foundations of Temporal Answer Set Programming through the lens of Temporal Equilibrium Logic, a formalism combining equilibrium logic with linear-time temporal operators. We lift the seminal approaches of Pearce and Osorio to the temporal setting, establishing a formal correspondence between temporal intuitionistic logic and temporal logic programming. Our results deepen the theoretical underpinnings of Temporal Answer Set Programming and provide new avenues for research in temporal reasoning.

Protocols for causal message delivery are widely used in distributed systems. Traditionally, causal delivery can be enforced either on the message sender's side or on the receiver's side. The traditional sender-side approach avoids the message metadata overhead of the receiver-side approach, but is more conservative than necessary. We present Cykas ("Can you keep a secret?"), a new protocol for sender-side enforcement of causal delivery that sidesteps the conservativeness of the traditional sender-side approach by allowing eager sending of messages and constraining the behavior of their recipients. We implemented the Cykas protocol in Rust and checked the safety and liveness of our implementation using the Stateright implementation-level model checker. Our experiments show that for applications involving long-running jobs, Cykas has a performance advantage: Cykas lets long-running jobs start (and end) earlier, leading to shorter overall execution time compared to the traditional sender-side approach.

As multi-agent AI systems are increasingly deployed in real-world settings - from automated customer support to DevOps remediation - failures become harder to diagnose due to cascading effects, hidden dependencies, and long execution traces. We present AgentTrace, a lightweight causal tracing framework for post-hoc failure diagnosis in deployed multi-agent workflows. AgentTrace reconstructs causal graphs from execution logs, traces backward from error manifestations, and ranks candidate root causes using interpretable structural and positional signals - without requiring LLM inference at debugging time. Across a diverse benchmark of multi-agent failure scenarios designed to reflect common deployment patterns, AgentTrace localizes root causes with high accuracy and sub-second latency, significantly outperforming both heuristic and LLM-based baselines. Our results suggest that causal tracing provides a practical foundation for improving the reliability and trustworthiness of agentic systems in the wild.

Human-Object Interaction (HOI) video reenactment with realistic motion remains a frontier in expressive digital human creation. Existing approaches primarily handle simple image-plane motion (e.g., in-plane translations), struggling with complex non-planar manipulations like out-of-plane reorientation. In this paper, we propose MVHOI, a two-stage HOI video reenactment framework that bridges multi-view reference conditions and video foundation models via a 3D Foundation Model (3DFM). The 3DFM first produces view-consistent object priors conditioned on implicit motion dynamics across novel viewpoints. A controllable video generation model then synthesizes high-fidelity object texture by incorporating multi-view reference images, ensuring appearance consistency via a reasonable retrieval mechanism. By enabling these two stages to mutually reinforce one another during the inference phase, our framework shows superior performance in generating long-duration HOI videos with intricate object manipulations. Extensive experiments show substantial improvements over prior approaches, especially for HOI with complex 3D object manipulations.

Bayesian change-point and segmentation models provide uncertainty-aware piecewise-constant representations of ordered data, but exact inference is often tied to narrow likelihood classes, single-sequence settings, or index-uniform designs. We present \texttt{BayesBreak}, a modular offline Bayesian segmentation framework built around a simple separation: each candidate block contributes a marginal likelihood and any required moment numerators, and a global dynamic program combines those block scores into posterior quantities over segment counts, boundary locations, and latent signals. For weighted exponential-family likelihoods with conjugate priors, block evidences and posterior moments are available in closed form from cumulative sufficient statistics, yielding exact sum-product inference for $P(y\mid k)$, $P(k\mid y)$, boundary marginals, and Bayes regression curves. We also distinguish these quantities from the \emph{joint} MAP segmentation, which is recovered by a separate max-sum backtracking recursion.

This study investigates the potential influence of Herman Melville reading on his own writings through computational semantic similarity analysis. Using documented records of books known to have been owned or read by Melville, we compare selected passages from his works with texts from his library. The methodology involves segmenting texts at both sentence level and non-overlapping 5-gram level, followed by similarity computation using BERTScore. Rather than applying fixed thresholds to determine reuse, we interpret precision, recall, and F1 scores as indicators of possible semantic alignment that may suggest literary influence. Experimental results demonstrate that the approach successfully captures expert-identified instances of similarity and highlights additional passages warranting further qualitative examination. The findings suggest that semantic similarity methods provide a useful computational framework for supporting source and influence studies in literary scholarship.

We study nonstationary Online Linear Programming (OLP), where $n$ orders arrive sequentially with reward-resource consumption pairs that form a sequence of independent, but not necessarily identically distributed, random vectors. At the beginning of the planning horizon, the decision-maker is provided with a resource endowment that is sufficient to fulfill a significant portion of the requests. The decision-maker seeks to maximize the expected total reward by making immediate and irrevocable acceptance or rejection decisions for each order, subject to this resource endowment. We focus on the challenging single-sample setting, where only one sample from each of the $n$ distributions is available at the start of the planning horizon. We propose a novel re-solving algorithm that integrates a dynamic programming perspective with the dual-based frameworks traditionally employed in stationary environments. In the large-resource regime, where the resource endowment scales linearly with the number of orders, we prove that our algorithm achieves $O((\log n)^2)$ regret across a broad class of nonstationary distribution sequences. Our results demonstrate that polylogarithmic regret is attainable even under significant environmental shifts and minimal data availability, bridging the gap between stationary OLP and more volatile real-world resource allocation problems.

Language Models (LMs) may acquire harmful knowledge, and yet feign ignorance of these topics when under audit. Inspired by the recent discovery of deception-related behaviour patterns in LMs, we aim to train classifiers that detect when a LM is actively concealing knowledge. Initial findings on smaller models show that classifiers can detect concealment more reliably than human evaluators, with gradient-based concealment proving easier to identify than prompt-based methods. However, contrary to prior work, we find that the classifiers do not reliably generalize to unseen model architectures and topics of hidden knowledge. Most concerningly, the identifiable traces associated with concealment become fainter as the models increase in scale, with the classifiers achieving no better than random performance on any model exceeding 70 billion parameters. Our results expose a key limitation in black-box-only auditing of LMs and highlight the need to develop robust methods to detect models that are actively hiding the knowledge they contain.

Embodied reasoning is inherently viewpoint-dependent: what is visible, occluded, or reachable depends critically on where the agent stands. However, existing spatial memory systems for embodied agents typically store either multi-view observations or object-centric abstractions, making it difficult to perform reasoning with explicit geometric grounding. We introduce RenderMem, a spatial memory framework that treats rendering as the interface between 3D world representations and spatial reasoning. Instead of storing fixed observations, RenderMem maintains a 3D scene representation and generates query-conditioned visual evidence by rendering the scene from viewpoints implied by the query. This enables embodied agents to reason directly about line-of-sight, visibility, and occlusion from arbitrary perspectives. RenderMem is fully compatible with existing vision-language models and requires no modification to standard architectures. Experiments in the AI2-THOR environment show consistent improvements on viewpoint-dependent visibility and occlusion queries over prior memory baselines.

Magnetic resonance imaging (MRI) super-resolution (SR) methods that computationally enhance low-resolution acquisitions to approximate high-resolution quality offer a compelling alternative to expensive high-field scanners. In this work we investigate an elucidated diffusion model (EDM) framework for brain MRI SR and compare two U-Net backbone architectures: (i) a full 3D convolutional U-Net that processes volumetric patches with 3D convolutions and multi-head self-attention, and (ii) a 2.5D slice-conditioned U-Net that super-resolves each slice independently while conditioning on an adjacent slice for inter-slice context. Both models employ continuous-sigma noise conditioning following Karras et al. and are trained on the NKI cohort of the FOMO60K dataset. On a held-out test set of 5 subjects (6 volumes, 993 slices), the 3D model achieves 37.75 dB PSNR, 0.997 SSIM, and 0.020 LPIPS, improving on the off-the-shelf pretrained EDSR baseline (35.57 dB / 0.024 LPIPS) and the 2.5D variant (35.82 dB) across all three metrics under the same test data and degradation pipeline.

Training data attribution (TDA) methods ask which training documents are responsible for a model behavior. We argue that this per-document framing is fundamentally mismatched to how fine-tuning actually works: models often learn broad concepts shared across many examples. Existing TDA methods are supervised -- they require a query behavior, then score every training document against it -- making them both expensive and unable to surface behaviors the user did not think to ask about. We present Gradient Atoms, an unsupervised method that decomposes per-document training gradients into sparse components ("atoms") via dictionary learning in a preconditioned eigenspace. Among the 500 discovered atoms, the highest-coherence ones recover interpretable task-type behaviors -- refusal, arithmetic, yes/no classification, trivia QA -- without any behavioral labels. These atoms double as effective steering vectors: applying them as weight-space perturbations produces large, controllable shifts in model behavior (e.g., bulleted-list generation 33% to 94%; systematic refusal 50% to 0%). The method requires no query--document scoring stage, and scales independently of the number of query behaviors of interest. Code is here: https://github.com/jrosseruk/gradient_atoms

The dominant narrative of artificial intelligence development assumes that progress is continuous and that capability scales monotonically with model size. We challenge both assumptions. Drawing on punctuated equilibrium theory from evolutionary biology, we show that AI development proceeds not through smooth advancement but through extended periods of stasis interrupted by rapid phase transitions that reorganize the competitive landscape. We identify five such eras since 1943 and four epochs within the current Generative AI Era, each initiated by a discontinuous event -- from the transformer architecture to the DeepSeek Moment -- that rendered the prior paradigm subordinate. To formalize the selection pressures driving these transitions, we develop the Institutional Fitness Manifold, a mathematical framework that evaluates AI systems along four dimensions: capability, institutional trust, affordability, and sovereign compliance. The central result is the Institutional Scaling Law, which proves that institutional fitness is non-monotonic in model scale. Beyond an environment-specific optimum, scaling further degrades fitness as trust erosion and cost penalties outweigh marginal capability gains. This directly contradicts classical scaling laws and carries a strong implication: orchestrated systems of smaller, domain-adapted models can mathematically outperform frontier generalists in most institutional deployment environments. We derive formal conditions under which this inversion holds and present supporting empirical evidence spanning frontier laboratory dynamics, post-training alignment evolution, and the rise of sovereign AI as a geopolitical selection pressure.

Video reasoning requires models to locate and track question-relevant evidence across frames. While reinforcement learning (RL) with verifiable rewards improves accuracy, it still struggles to achieve reliable spatio-temporal grounding during the reasoning process. Moreover, improving grounding typically relies on scaled training data or inference-time perception tools, which increases annotation cost or computational cost. To address this challenge, we propose VisonCoach, an input-adaptive RL framework that improves spatio-temporal grounding through visual prompting as training-time guidance. During RL training, visual prompts are selectively applied to challenging inputs to amplify question-relevant evidence and suppress distractors. The model then internalizes these improvements through self-distillation, enabling grounded reasoning directly on raw videos without visual prompting at inference. VisonCoach consists of two components: (1) Visual Prompt Selector, which predicts appropriate prompt types conditioned on the video and question, and (2) Spatio-Temporal Reasoner, optimized with RL under visual prompt guidance and object-aware grounding rewards that enforce object identity consistency and multi-region bounding-box overlap. Extensive experiments demonstrate that VisonCoach achieves state-of-the-art performance under comparable settings, across diverse video reasoning, video understanding, and temporal grounding benchmarks (V-STAR, VideoMME, World-Sense, VideoMMMU, PerceptionTest, and Charades-STA), while maintaining a single efficient inference pathway without external tools. Our results show that visual prompting during training improves grounded video reasoning, while self-distillation enables the model to internalize this ability without requiring prompts at inference time.

Deepfake detection is widely framed as a machine learning problem, yet how humans and AI detectors compare under realistic conditions remains poorly understood. We evaluate 200 human participants and 95 state-of-the-art AI detectors across two datasets: DF40, a standard benchmark, and CharadesDF, a novel dataset of videos of everyday activities. CharadesDF was recorded using mobile phones leading to low/moderate quality videos compared to the more professionally captured DF40. Humans outperform AI detectors on both datasets, with the gap widening in the case of CharadesDF where AI accuracy collapses to near chance (0.537) while humans maintain robust performance (0.784). Human and AI errors are complementary: humans miss high-quality deepfakes while AI detectors flag authentic videos as fake, and hybrid human-AI ensembles reduce high-confidence errors. These findings suggest that effective real-world deepfake detection, especially in non-professionally produced videos, requires human-AI collaboration rather than AI algorithms alone.

We present EARCP (Ensemble Auto-Régulé par Cohérence et Performance), a novel ensemble architecture that dynamically weights heterogeneous expert models based on both their individual performance and inter-model coherence. Unlike traditional ensemble methods that rely on static or offline-learned combinations, EARCP continuously adapts model weights through a principled online learning mechanism that balances exploitation of high-performing models with exploration guided by consensus signals. The architecture combines theoretical foundations from multiplicative weight update algorithms with a novel coherence-based regularization term, providing both theoretical guarantees through regret bounds and practical robustness in non-stationary environments. We formalize the EARCP framework, prove sublinear regret bounds of O(sqrt(T log M)) under standard assumptions, and demonstrate its effectiveness through empirical evaluation on sequential prediction tasks including time series forecasting, activity recognition, and financial prediction. The architecture is designed as a general-purpose framework applicable to any domain requiring ensemble learning with temporal dependencies. An open-source implementation is available at https://github.com/Volgat/earcp and via PyPI (pip install earcp).

Nuclear power plant operators face significant challenges due to unpredictable deviations between offline and online thermal limits, a phenomenon known as thermal limit bias, which leads to conservative design margins, increased fuel costs, and operational inefficiencies. This work presents a deep learning based methodology to predict and correct this bias for Boiling Water Reactors (BWRs), focusing on the Maximum Fraction of Limiting Power Density (MFLPD) metric used to track the Linear Heat Generation Rate (LHGR) limit. The proposed model employs a fully convolutional encoder decoder architecture, incorporating a feature fusion network to predict corrected MFLPD values closer to online measurements. Evaluated across five independent fuel cycles, the model reduces the mean nodal array error by 74 percent, the mean absolute deviation in limiting values by 72 percent, and the maximum bias by 52 percent compared to offline methods. These results demonstrate the model's potential to meaningfully improve fuel cycle economics and operational planning, and a commercial variant has been deployed at multiple operating BWRs.

Contrastive learning (CL) has become a powerful approach for learning representations from unlabeled images. However, existing CL methods focus predominantly on visual appearance features while neglecting topological characteristics (e.g., connectivity patterns, boundary configurations, cavity formations) that provide valuable cues for medical image analysis. To address this limitation, we propose a new topological CL framework (TopoCL) that explicitly exploits topological structures during contrastive learning for medical imaging. Specifically, we first introduce topology-aware augmentations that control topological perturbations using a relative bottleneck distance between persistence diagrams, preserving medically relevant topological properties while enabling controlled structural variations. We then design a Hierarchical Topology Encoder that captures topological features through self-attention and cross-attention mechanisms. Finally, we develop an adaptive mixture-of-experts (MoE) module to dynamically integrate visual and topological representations. TopoCL can be seamlessly integrated with existing CL methods. We evaluate TopoCL on five representative CL methods (SimCLR, MoCo-v3, BYOL, DINO, and Barlow Twins) and five diverse medical image classification datasets. The experimental results show that TopoCL achieves consistent improvements: an average gain of +3.26% in linear probe classification accuracy with strong statistical significance, verifying its effectiveness.

Theory of Mind (ToM) is central to social cognition and human-AI interaction, and Large Language Models (LLMs) have been used to help understand and represent ToM. However, most evaluations treat ToM as a static judgment at a single moment, primarily relying on tests of false beliefs. This overlooks a key dynamic dimension of ToM: the ability to represent, update, and retrieve others' beliefs over time. We investigate dynamic ToM as a temporally extended representational memory problem, asking whether LLMs can track belief trajectories across interactions rather than only inferring current beliefs. We introduce DToM-Track, an evaluation framework to investigate temporal belief reasoning in controlled multiturn conversations, testing the recall of beliefs held prior to an update, the inference of current beliefs, and the detection of belief change. Using LLMs as computational probes, we find a consistent asymmetry: models reliably infer an agent's current belief but struggle to maintain and retrieve prior belief states once updates occur. This pattern persists across LLM model families and scales, and is consistent with recency bias and interference effects well documented in cognitive science. These results suggest that tracking belief trajectories over time poses a distinct challenge beyond classical false-belief reasoning. By framing ToM as a problem of temporal representation and retrieval, this work connects ToM to core cognitive mechanisms of memory and interference and exposes the implications for LLM models of social reasoning in extended human-AI interactions.

Publicly available full-field digital mammography (FFDM) datasets remain limited in size, clinical labels, and vendor diversity, which hinders the training of robust models. We present LUMINA, a curated, multi-vendor FFDM dataset that explicitly encodes acquisition energy and vendor metadata to expose clinically relevant appearance shifts that current benchmarks overlook. This innovative resource comprises 1824 images from 468 patients (960 benign, 864 malignant) with pathology-confirmed outcomes, BI-RADS assessments, and breast-density annotations. LUMINA spans six acquisition systems and both high- and low-energy styles, exposing vendor- and energy-driven appearance shifts. To reduce cross-vendor/energy drift while preserving lesion morphology, we introduce a foreground-only, pixel-space alignment (''energy harmonization'') that aligns each image to a low-energy reference style, leaving the zero-valued background unchanged. By benchmarking modern CNN and transformer baselines on three clinically meaningful tasks -- diagnosis (benign vs. malignant), BI-RADS risk grouping, and density -- we unify single-vs-two-view evaluation and show that two-view models consistently outperform single-view; in our benchmark, EfficientNet-B0 attains AUC 93.54% for diagnosis, and Swin-T yields the best macro-AUC 89.43% for density. Harmonization improves AUC/ACC across backbones and yields more focal Grad-CAM localization around suspicious regions. Being a richly annotated resource, LUMINA thus provides (a) a vendor-diverse, energy-labeled benchmark and (b) a model-agnostic harmonization protocol that together catalyze reliable, deployable mammography AI.

Large language models (LLMs) exhibit strong general capabilities, but their deployment in high-stakes domains is hindered by their opacity and unpredictability. Recent work has taken meaningful steps towards addressing these issues by augmenting LLMs with post-hoc reasoning based on computational argumentation, providing faithful explanations and enabling users to contest incorrect decisions. However, this paradigm is limited to pre-defined binary choices and only supports local contestation for specific instances, leaving the underlying decision logic unchanged and prone to repeated mistakes. In this paper, we introduce ArgEval, a framework that shifts from instance-specific reasoning to structured evaluation of general decision options. Rather than mining arguments solely for individual cases, ArgEval systematically maps task-specific decision spaces, builds corresponding option ontologies, and constructs general argumentation frameworks (AFs) for each option. These frameworks can then be instantiated to provide explainable recommendations for specific cases while still supporting global contestability through modification of the shared AFs. We investigate the effectiveness of ArgEval on treatment recommendation for glioblastoma, an aggressive brain tumour, and show that it can produce explainable guidance aligned with clinical practice.

Chain-of-thought (CoT) prompting has been extended to large audio-language models (LALMs) to elicit reasoning, yet enhancing its effectiveness without training remains challenging. We study inference-time model steering as a training-free approach to improve LALM reasoning. We introduce three strategies using diverse information sources and evaluate them across four LALMs and four benchmarks. Results show general accuracy gains up to 4.4% over CoT prompting. Notably, we identify a cross-modal transfer where steering vectors derived from few text samples effectively guide speech-based reasoning, demonstrating high data efficiency. We also examine hyperparameter sensitivity to understand the robustness of these approaches. Our findings position model steering as a practical direction for strengthening LALM reasoning.

As agents operate over long horizons, their memory stores grow continuously, making retrieval critical to accessing relevant information. Many agent queries require reasoning-intensive retrieval, where the connection between query and relevant documents is implicit and requires inference to bridge. LLM-augmented pipelines address this through query expansion and candidate re-ranking, but introduce significant inference costs. We study computation allocation in reasoning-intensive retrieval pipelines using the BRIGHT benchmark and Gemini 2.5 model family. We vary model capacity, inference-time thinking, and re-ranking depth across query expansion and re-ranking stages. We find that re-ranking benefits substantially from stronger models (+7.5 NDCG@10) and deeper candidate pools (+21% from $k$=10 to 100), while query expansion shows diminishing returns beyond lightweight models (+1.1 NDCG@10 from weak to strong). Inference-time thinking provides minimal improvement at either stage. These results suggest that compute should be concentrated on re-ranking rather than distributed uniformly across pipeline stages.

Increasing staffing constraints and turnaround-time pressures in Prior authorization (PA) have led to increasing automation of decision systems to support PA review. Evaluating fairness in such systems poses unique challenges because legitimate clinical guidelines and medical necessity criteria often differ across demographic groups, making parity in approval rates an inappropriate fairness metric. We propose a fairness evaluation framework for prior authorization models based on model error rates rather than approval outcomes. Using 7,166 human-reviewed cases spanning 27 medical necessity guidelines, we assessed consistency in sex, age, race/ethnicity, and socioeconomic status. Our evaluation combined error-rate comparisons, tolerance-band analysis with a predefined $\pm$5 percentage-point margin, statistical power evaluation, and protocol-controlled logistic regression. Across most demographics, model error rates were consistent, and confidence intervals fell within the predefined tolerance band, indicating no meaningful performance differences. For race/ethnicity, point estimates remain small, but subgroup sample sizes were limited, resulting in wide confidence intervals and underpowered tests, with inconclusive evidence within the dataset we explored. These findings illustrate a rigorous and regulator-aligned approach to fairness evaluation in administrative healthcare AI systems.

The ongoing convergence of HPC and cloud computing presents a fundamental challenge: HPC applications, designed for static and homogeneous supercomputers, are ill-suited for the dynamic, heterogeneous, and volatile nature of the cloud. Traditional parallel programming models like MPI struggle to leverage key cloud advantages, such as resource elasticity and low-cost spot instances, while also failing to address challenges like performance variability and processor heterogeneity. This paper demonstrates how the asynchronous, message-driven paradigm of the Charm++ parallel runtime system can bridge this gap. We present a set of tools and strategies that enable HPC applications to run efficiently and resiliently on dynamic cloud infrastructure across both CPU and GPU resources. Our work makes two key contributions. First, we demonstrate that rate-aware load balancing in Charm++ improves performance for applications running on heterogeneous CPU and GPU instances on the cloud. We further demonstrate how core Charm++ principles mitigate performance degradation from common cloud challenges like network contention and processor performance variability, which are exacerbated by the tightly coupled, globally synchronized nature of many science and engineering applications. Second, we extend an existing resource management framework to support GPU and CPU spot instances with minimal interruption overhead. Together, these contributions provide a robust framework for adapting HPC applications to achieve efficient, resilient, and cost-effective performance on the cloud.

ResearchPilot is an open-source, self-hostable multi-agent system for literature-review assistance. Given a natural-language research question, it retrieves papers from Semantic Scholar and arXiv, extracts structured findings from paper abstracts, synthesizes cross-paper patterns, and drafts a citation-aware related-work section. The system combines FastAPI, Next.js, DSPy, SQLite, and Qdrant in a local-first architecture that supports bring-your-own-key model access and remote-or-local embeddings. This paper describes the system design, typed agent interfaces, persistence and history-search mechanisms, and the engineering tradeoffs involved in building a transparent research assistant. Rather than claiming algorithmic novelty, we present ResearchPilot as a systems contribution and evaluate it through automated tests and end-to-end local runs. We discuss limitations including external API rate limits, abstract-only extraction, incomplete corpus coverage, and the lack of citation verification.

Neurosymbolic approaches leveraging Large Language Models (LLMs) with formal methods have recently achieved strong results on mathematics-oriented theorem-proving benchmarks. However, success on competition-style mathematics does not by itself demonstrate the ability to construct proofs about real-world implementations. We address this gap with a benchmark derived from an industrial cryptographic library whose assembly routines are already verified in HOL Light. s2n-bignum is a library used at AWS for providing fast assembly routines for cryptography, and its correctness is established by formal verification. The task of formally verifying this library has been a significant achievement for the Automated Reasoning Group. It involved two tasks: (1) precisely specifying the correct behavior of a program as a mathematical proposition, and (2) proving that the proposition is correct. In the case of s2n-bignum, both tasks were carried out by human experts. In \textit{s2n-bignum-bench}, we provide the formal specification and ask the LLM to generate a proof script that is accepted by HOL Light within a fixed proof-check timeout. To our knowledge, \textit{s2n-bignum-bench} is the first public benchmark focused on machine-checkable proof synthesis for industrial low-level cryptographic assembly routines in HOL Light. This benchmark provides a challenging and practically relevant testbed for evaluating LLM-based theorem proving beyond competition mathematics. The code to set up and use the benchmark is available here: \href{https://github.com/kings-crown/s2n-bignum-bench}{s2n-bignum-bench}.

Global decarbonisation targets and tightening market pressures demand maritime logistics solutions that are simultaneously efficient, sustainable, and equitable. We introduce EcoFair-CH-MARL, a constrained hierarchical multi-agent reinforcement learning framework that unifies three innovations: (i) a primal-dual budget layer that provably bounds cumulative emissions under stochastic weather and demand; (ii) a fairness-aware reward transformer with dynamically scheduled penalties that enforces max-min cost equity across heterogeneous fleets; and (iii) a two-tier policy architecture that decouples strategic routing from real-time vessel control, enabling linear scaling in agent count. New theoretical results establish O(\sqrt{T}) regret for both constraint violations and fairness loss. Experiments on a high-fidelity maritime digital twin (16 ports, 50 vessels) driven by automatic identification system traces, plus an energy-grid case study, show up to 15% lower emissions, 12% higher through-put, and a 45% fair-cost improvement over state-of-the-art hierarchical and constrained MARL baselines. In addition, EcoFair-CH-MARL achieves stronger equity (lower Gini and higher min-max welfare) than fairness-specific MARL baselines (e.g., SOTO, FEN), and its modular design is compatible with both policy- and value-based learners. EcoFair-CH-MARL therefore advances the feasibility of large-scale, regulation-compliant, and socially responsible multi-agent coordination in safety-critical domains.

Model routing determines whether to use an accurate black-box model or a simpler surrogate that approximates it at lower cost or greater interpretability. In deployment settings, practitioners often wish to restrict surrogate use to inputs where its degradation relative to a reference model is controlled. We study proactive (input-based) routing, in which a lightweight gate selects the model before either runs, enabling distribution-free control of the fraction of routed inputs whose degradation exceeds a tolerance τ. The gate is trained to distinguish safe from unsafe inputs, and a routing threshold is chosen via Clopper-Pearson conformal calibration on a held-out set, guaranteeing that the routed-set violation rate is at most α with probability 1-δ. We derive a feasibility condition linking safe routing to the base safe rate π and risk budget α, along with sufficient AUC thresholds ensuring that feasible routing exists. Across 35 OpenML datasets and multiple black-box model families, gate-based conformal routing maintains controlled violation while achieving substantially higher coverage than regression conformal and naive baselines. We further show that probabilistic calibration primarily affects routing efficiency rather than distribution-free validity.

Cloud-native software delivery platforms orchestrate releases through complex, multi-stage pipelines composed of dozens of independently versioned tasks. When code is promoted between environments -- development to staging, staging to production -- engineering teams need timely, accurate communication about what changed and what downstream components are affected. Manual preparation of such release communication is slow, inconsistent, and particularly error-prone in repositories where a single promotion may bundle contributions from many authors across numerous pipeline tasks. We present a framework for AI-augmented release intelligence that combines three capabilities: (1) automated commit collection with semantic filtering to surface substantive changes while suppressing routine maintenance, (2) structured large language model summarization that produces categorized, stakeholder-oriented promotion reports, and (3) static task-pipeline dependency analysis that maps modified tasks to every pipeline they participate in, quantifying the blast radius of each change. The framework is integrated directly into the CI/CD promotion workflow and operates as a post-promotion step triggered by GitHub Actions. We describe the architecture and implementation within a production Kubernetes-native release platform that manages over sixty Tekton tasks across more than twenty release pipelines. Through concrete walkthrough examples and qualitative comparison with recent tools such as SmartNote and VerLog, we discuss the distinctive requirements of internal promotion communication versus user-facing release notes and identify open challenges for LLM-driven release engineering.

This paper presents the functional safety analysis for an Infrastructure-Enabled Depot Autonomy (IX-DA) system. The IX-DA system automates the marshalling of delivery vehicles within a controlled depot environment, navigating connected autonomous vehicles (CAVs) between drop-off zones, service stations (washing, calibration, charging, loading), and pick-up zones without human intervention. We describe the system architecture comprising three principal subsystems -- the connected autonomous vehicle, the infrastructure sensing and compute layer, and the human operator interface -- and derive their functional requirements. Using ISO 26262-compliant Hazard Analysis and Risk Assessment (HARA) methodology, we identify eight hazardous events, evaluate them across different operating scenarios, and assign Automotive Safety Integrity Levels~(ASILs) ranging from Quality Management (QM) to ASIL C. Six safety goals are derived and allocated to vehicle and infrastructure subsystems. The analysis demonstrates that high-speed uncontrolled operation imposes the most demanding safety requirements (ASIL C), while controlled low-speed operation reduces most goals to QM, offering a practical pathway for phased deployment.

Standard policy gradients weight each sampled action by advantage alone, regardless of how likely that action was under the current policy. This creates two pathologies: within a single decision context (e.g. one image or prompt), a rare negative-advantage action can disproportionately distort the update direction; across many such contexts in a batch, the expected gradient over-allocates budget to contexts the policy already handles well. We introduce the \textit{Delightful Policy Gradient} (DG), which gates each term with a sigmoid of \emph{delight}, the product of advantage and action surprisal (negative log-probability). For $K$-armed bandits, DG provably improves directional accuracy in a single context and, across multiple contexts, shifts the expected gradient strictly closer to the supervised cross-entropy oracle. This second effect is not variance reduction: it persists even with infinite samples. Empirically, DG outperforms REINFORCE, PPO, and advantage-weighted baselines across MNIST, transformer sequence modeling, and continuous control, with larger gains on harder tasks.

We propose TacFiLM, a lightweight modality-fusion approach that integrates visual-tactile signals into vision-language-action (VLA) models. While recent advances in VLA models have introduced robot policies that are both generalizable and semantically grounded, these models mainly rely on vision-based perception. Vision alone, however, cannot capture the complex interaction dynamics that occur during contact-rich manipulation, including contact forces, surface friction, compliance, and shear. While recent attempts to integrate tactile signals into VLA models often increase complexity through token concatenation or large-scale pretraining, the heavy computational demands of behavioural models necessitate more lightweight fusion strategies. To address these challenges, TacFiLM outlines a post-training finetuning approach that conditions intermediate visual features on pretrained tactile representations using feature-wise linear modulation (FiLM). Experimental results on insertion tasks demonstrate consistent improvements in success rate, direct insertion performance, completion time, and force stability across both in-distribution and out-of-distribution tasks. Together, these results support our method as an effective approach to integrating tactile signals into VLA models, improving contact-rich manipulation behaviours.

Conversational assistants powered by large language models (LLMs) excel at tool-use tasks but struggle with adhering to complex, business-specific rules. While models can reason over business rules provided in context, including all policies for every query introduces high latency and wastes compute. Furthermore, these lengthy prompts lead to long contexts, harming overall performance due to the "needle-in-the-haystack" problem. To address these challenges, we propose a multi-stage alignment method that teaches models to recall and apply relevant business policies during chain-of-thought reasoning at inference time, without including the full business policy in-context. Furthermore, we introduce a novel PolicyRecall reward based on the Jaccard score and a Hallucination Penalty for GRPO training. Altogether, our best model outperforms the baseline by 16 points and surpasses comparable in-context baselines of similar model size by 3 points, while using 40% fewer words.

We study the Fréchet {\it k-}means of a metric measure space when both the measure and the distance are unknown and have to be estimated. We prove a general result that states that the {\it k-}means are continuous with respect to the measured Gromov-Hausdorff topology. In this situation, we also prove a stability result for the Voronoi clusters they determine. We do not assume uniqueness of the set of {\it k-}means, but when it is unique, the results are stronger. {This framework provides a unified approach to proving consistency for a wide range of metric learning procedures. As concrete applications, we obtain new consistency results for several important estimators that were previously unestablished, even when $k=1$. These include {\it k-}means based on: (i) Isomap and Fermat geodesic distances on manifolds, (ii) difussion distances, (iii) Wasserstein distances computed with respect to learned ground metrics. Finally, we consider applications beyond the statistical inference paradigm like (iv) first passage percolation and (v) discrete approximations of length spaces.}

Reinforcement learning algorithms have been widely used in dynamic and control systems. However, interpreting their internal learning behavior remains a challenge. In the authors' previous work, a critic match loss landscape visualization method was proposed to study critic training. This study extends that method into a framework which provides a multi-perspective view of the learning dynamics, clarifying how value estimation, policy optimization, and temporal-difference (TD) signals interact during training. The proposed framework includes four complementary components; a three-dimensional reconstruction of the critic match loss surface that shows how TD targets shape the optimization geometry; an actor loss landscape under a frozen critic that reveals how the policy exploits that geometry; a trajectory combining time, Bellman error, and policy weights that indicates how updates move across the surface; and a state-TD map that identifies the state regions that drive those updates. The Action-Dependent Heuristic Dynamic Programming (ADHDP) algorithm for spacecraft attitude control is used as a case study. The framework is applied to compare several ADHDP variants and shows how training stabilizers and target updates change the optimization landscape and affect learning stability. Therefore, the proposed framework provides a systematic and interpretable tool for analyzing reinforcement learning behavior across algorithmic designs.

Autonomous LLM agents require structured long-term memory, yet current "append-and-evolve" systems like A-MEM face O(N^2) write-latency and excessive token costs. We introduce D-MEM (Dopamine-Gated Agentic Memory), a biologically inspired architecture that decouples short-term interaction from cognitive restructuring via a Fast/Slow routing system based on Reward Prediction Error (RPE). A lightweight Critic Router evaluates stimuli for Surprise and Utility. Routine, low-RPE inputs are bypassed or cached in an O(1) fast-access buffer. Conversely, high-RPE inputs, such as factual contradictions or preference shifts, trigger a "dopamine" signal, activating the O(N) memory evolution pipeline to reshape the agent's knowledge graph. To evaluate performance under realistic conditions, we introduce the LoCoMo-Noise benchmark, which injects controlled conversational noise into long-term sessions. Evaluations demonstrate that D-MEM reduces token consumption by over 80%, eliminates O(N^2) bottlenecks, and outperforms baselines in multi-hop reasoning and adversarial resilience. By selectively gating cognitive restructuring, D-MEM provides a scalable, cost-efficient foundation for lifelong agentic memory.

We propose a new algorithm for compiling Bayesian network classifier (BNC) into class formulas. Class formulas are logical formulas that represent a classifier's input-output behavior, and are crucial in the recent line of work that uses logical reasoning to explain the decisions made by classifiers. Compared to prior work on compiling class formulas of BNCs, our proposed algorithm is not restricted to binary classifiers, shows significant improvement in compilation time, and outputs class formulas as negation normal form (NNF) circuits that are OR-decomposable, which is an important property when computing explanations of classifiers.

Tiny Recursive Models (TRM) achieve strong results on reasoning tasks through iterative refinement of a shared network. We investigate whether these recursive mechanisms transfer to Quality Estimation (QE) for low-resource languages using a three-phase methodology. Experiments on $8$ language pairs on a low-resource QE dataset reveal three findings. First, TRM's recursive mechanisms do not transfer to QE. External iteration hurts performance, and internal recursion offers only narrow benefits. Next, representation quality dominates architectural choices, and lastly, frozen pretrained embeddings match fine-tuned performance while reducing trainable parameters by 37$\times$ (7M vs 262M). TRM-QE with frozen XLM-R embeddings achieves a Spearman's correlation of 0.370, matching fine-tuned variants (0.369) and outperforming an equivalent-depth standard transformer (0.336). On Hindi and Tamil, frozen TRM-QE outperforms MonoTransQuest (560M parameters) with 80$\times$ fewer trainable parameters, suggesting that weight sharing combined with frozen embeddings enables parameter efficiency for QE. We release the code publicly for further research. Code is available at https://github.com/surrey-nlp/TRMQE.

Financial fraud detection in transaction networks involves modeling sparse anomalies, dynamic patterns, and severe class imbalance in the presence of temporal drift in the data. In real-world transaction systems, a suspicious transaction is rarely isolated: rather, legitimate and suspicious transactions are often connected through accounts, intermediaries or through temporal transaction sequences. Attribute-based or randomly partitioned learning pipelines are therefore insufficient to detect relationally structured fraud. STC-MixHop, a graph-based framework combining spatial multi-resolution propagation with lightweight temporal consistency modeling for anomaly and fraud detection in dynamic transaction networks. It integrates three components: a MixHop-inspired multi-scale neighborhood diffusion encoder a multi-scale neighborhood diffusion MixHop-based encoder for learning structural patterns; a spatial-temporal attention module coupling current and preceding graph snapshots to stabilize representations; and a temporally informed self-supervised pretraining strategy exploiting unlabeled transaction interactions to improve representation quality. We evaluate the framework primarily on the PaySim dataset under strict chronological splits, supplementing the analysis with Porto Seguro and FEMA data to probe cross-domain component behavior. Results show that STC-MixHop is competitive among graph methods and achieves strong screening-oriented recall under highly imbalanced conditions. The experiments also reveal an important boundary condition: when node attributes are highly informative, tabular baselines remain difficult to outperform. Graph structure contributes most clearly where hidden relational dependencies are operationally important. These findings support a stability-focused view of graph learning for financial fraud detection.

Language models are increasingly adopting smaller architectures optimized for consumer devices. In this setting, inference efficiency is the primary constraint. Meanwhile, vocabulary sizes continue to grow rapidly, making the classification head a critical bottleneck that accounts for up to 60\% of model parameters, and 50\% of inference compute. We introduce FlashHead, the first efficient drop-in replacement for the dense classification head that is training-free and hardware-friendly. FlashHead builds on principles from information retrieval, reframing that computation at the output head as a retrieval problem rather than a dense classification over the full vocabulary. FlashHead introduces four key innovations: (1) a balanced clustering scheme that structures vocabulary partitions into compact hardware-efficient tensors, (2) extending multiprobe retrieval to language model heads, enabling thousands of clusters to be scored in parallel, (3) a novel inference-time sampling mechanism that extends retrieval beyond top tokens, enabling probabilistic sampling across the full vocabulary, and (4) selective quantization, enabling effective low-bit computation in the head. Experiments on Llama-3.2, Gemma-3, and Qwen-3 show that FlashHead delivers model-level inference speedups of up to \textbf{1.75x} which maintaining output accuracy compared to the original head. By overcoming the classification head bottleneck, FlashHead establishes a new benchmark for efficient inference and removes a key barrier to developing smaller, capable models for consumer hardware.

This work extends an established critic match loss landscape visualization method from online to off-policy reinforcement learning (RL), aiming to reveal the optimization geometry behind critic learning. Off-policy RL differs from stepwise online actor-critic learning in its replay-based data flow and target computation. Based on these two structural differences, the critic match loss landscape visualization method is adapted to the Soft Actor-Critic (SAC) algorithm by aligning the loss evaluation with its batch-based data flow and target computation, using a fixed replay batch and precomputed critic targets from the selected policy. Critic parameters recorded during training are projected onto a principal component plane, where the critic match loss is evaluated to form a 3-D landscape with an overlaid 2-D optimization path. Applied to a spacecraft attitude control problem, the resulting landscapes are analyzed both qualitatively and quantitatively using sharpness, basin area, and local anisotropy metrics, together with temporal landscape snapshots. Comparisons between convergent SAC, divergent SAC, and divergent Action-Dependent Heuristic Dynamic Programming (ADHDP) cases reveal distinct geometric patterns and optimization behaviors under different algorithmic structures. The results demonstrate that the adapted critic match loss visualization framework serves as a geometric diagnostic tool for analyzing critic optimization dynamics in replay-based off-policy RL-based control problems.

Persistent memory is a central capability for AI agents, yet the mathematical foundations of memory retrieval, lifecycle management, and consistency remain unexplored. Current systems employ cosine similarity for retrieval, heuristic decay for salience, and provide no formal contradiction detection. We establish information-geometric foundations through three contributions. First, a retrieval metric derived from the Fisher information structure of diagonal Gaussian families, satisfying Riemannian metric axioms, invariant under sufficient statistics, and computable in O(d) time. Second, memory lifecycle formulated as Riemannian Langevin dynamics with proven existence and uniqueness of the stationary distribution via the Fokker-Planck equation, replacing hand-tuned decay with principled convergence guarantees. Third, a cellular sheaf model where non-trivial first cohomology classes correspond precisely to irreconcilable contradictions across memory contexts. On the LoCoMo benchmark, the mathematical layers yield +12.7 percentage points over engineering baselines across six conversations, reaching +19.9 pp on the most challenging dialogues. A four-channel retrieval architecture achieves 75% accuracy without cloud dependency. Cloud-augmented results reach 87.7%. A zero-LLM configuration satisfies EU AI Act data sovereignty requirements by architectural design. To our knowledge, this is the first work establishing information-geometric, sheaf-theoretic, and stochastic-dynamical foundations for AI agent memory systems.

As pedestrian navigation increasingly experiments with Generative AI, and in particular Large Language Models, the nature of routing risks transforming from a verifiable geometric task into an opaque, persuasive dialogue. While conversational interfaces promise personalisation, they introduce risks of manipulation and misplaced trust. We categorise these risks using a 2x2 framework based on intent and origin, distinguishing between intentional manipulations (dark patterns) and unintended harms (explainability pitfalls). We propose seamful design strategies to mitigate these harms. We suggest that one robust way to operationalise trustworthy conversational navigation is through neuro-symbolic architecture, where verifiable pathfinding algorithms ground GenAI's persuasive capabilities, ensuring systems explain their limitations and incentives as clearly as they explain the route.

Despite the data-rich environment in which memory systems of modern computing platforms operate, many state-of-the-art architectural policies employed in the memory system rely on static, human-designed heuristics that fail to truly adapt to the workload and system behavior via principled learning methodologies. In this article, we propose a fundamentally different design approach: using lightweight and practical machine learning (ML) methods to enable adaptive, data-driven control throughout the memory hierarchy. We present three ML-guided architectural policies: (1) Pythia, a reinforcement learning-based data prefetcher for on-chip caches, (2) Hermes, a perceptron learning-based off-chip predictor for multi-level cache hierarchies, and (3) Sibyl, a reinforcement learning-based data placement policy for hybrid storage systems. Our evaluation shows that Pythia, Hermes, and Sibyl significantly outperform the best-prior human-designed policies, while incurring modest hardware overheads. Collectively, this article demonstrates that integrating adaptive learning into memory subsystems can lead to intelligent, self-optimizing architectures that unlock performance and efficiency gains beyond what is possible with traditional human-designed approaches.

Vision language models (VLMs) have shown significant promise in visual grounding for images as well as videos. In medical imaging research, VLMs represent a bridge between object detection and segmentation, and report understanding and generation. However, spatial grounding of anatomical structures in the three-dimensional space of medical images poses many unique challenges. In this study, we examine image modalities, slice directions, and coordinate systems as differentiating factors for vision components of VLMs, and the use of anatomical, directional, and relational terminology as factors for the language components. We then demonstrate that visual and textual prompting systems such as labels, bounding boxes, and mask overlays have varying effects on the spatial grounding ability of VLMs. To enable measurement and reproducibility, we introduce \textbf{MIS-Ground}, a benchmark that comprehensively tests a VLM for vulnerabilities against specific modes of \textbf{M}edical \textbf{I}mage \textbf{S}patial \textbf{Ground}ing. We release MIS-Ground to the public at \href{https://anonymous.4open.science/r/mis-ground}{\texttt{anonymous.4open.science/r/mis-ground}}. In addition, we present \textbf{MIS-SemSam}, a low-cost, inference-time, and model-agnostic optimization of VLMs that improve their spatial grounding ability with the use of \textbf{Sem}antic \textbf{Sam}pling. We find that MIS-SemSam improves the accuracy of Qwen3-VL-32B on MIS-Ground by 13.06\%.

Scaling laws for neural networks, in which the loss decays as a power-law in the number of parameters, data, and compute, depend fundamentally on the spectral structure of the data covariance, with power-law eigenvalue decay appearing ubiquitously in vision and language tasks. A central question is whether this spectral structure is preserved or destroyed when data passes through the basic building block of a neural network: a random linear projection followed by a nonlinear activation. We study this question for the random feature model: given data $x \sim N(0,H)\in \mathbb{R}^v$ where $H$ has $α$-power-law spectrum ($λ_j(H ) \asymp j^{-α}$, $α> 1$), a Gaussian sketch matrix $W \in \mathbb{R}^{v\times d}$, and an entrywise monomial $f(y) = y^{p}$, we characterize the eigenvalues of the population random-feature covariance $\mathbb{E}_{x }[\frac{1}{d}f(W^\top x )^{\otimes 2}]$. We prove matching upper and lower bounds: for all $1 \leq j \leq c_1 d \log^{-(p+1)}(d)$, the $j$-th eigenvalue is of order $\left(\log^{p-1}(j+1)/j\right)^α$. For $ c_1 d \log^{-(p+1)}(d)\leq j\leq d$, the $j$-th eigenvalue is of order $j^{-α}$ up to a polylog factor. That is, the power-law exponent $α$ is inherited exactly from the input covariance, modified only by a logarithmic correction that depends on the monomial degree $p$. The proof combines a dyadic head-tail decomposition with Wick chaos expansions for higher-order monomials and random matrix concentration inequalities.

For deep learning inference on edge devices, hardware configurations achieving the same throughput can differ by 2$\times$ in power consumption, yet operators often struggle to find the efficient ones without exhaustive profiling. Existing approaches often rely on inefficient static presets or require expensive offline profiling that must be repeated for each new model or device. To address this problem, we present CORAL, an online optimization method that discovers near-optimal configurations without offline profiling. CORAL leverages distance covariance to statistically capture the non-linear dependencies between hardware settings, e.g., DVFS and concurrency levels, and performance metrics. Unlike prior work, we explicitly formulate the challenge as a throughput-power co-optimization problem to satisfy power budgets and throughput targets simultaneously. We evaluate CORAL on two NVIDIA Jetson devices across three object detection models ranging from lightweight to heavyweight. In single-target scenarios, CORAL achieves 96% $\unicode{x2013}$ 100% of the optimal performance found by exhaustive search. In strict dual-constraint scenarios where baselines fail or exceed power budgets, CORAL consistently finds proper configurations online with minimal exploration.

Quantum circuit Born machines based on instantaneous quantum polynomial-time (IQP) circuits are natural candidates for quantum generative modeling, both because of their probabilistic structure and because IQP sampling is provably classically hard in certain regimes. Recent proposals focus on training IQP-QCBMs using Maximum Mean Discrepancy (MMD) losses built from low-body Pauli-$Z$ correlators, but the effect of initialization on the resulting optimization landscape remains poorly understood. In this work, we address this by first proving that the MMD loss landscape suffers from barren plateaus for random full-angle-range initializations of IQP circuits. We then establish lower bounds on the loss variance for identity and an unbiased data-agnostic initialization. We then additionally consider a data-dependent initialization that is better aligned with the target distribution and, under suitable assumptions, yields provable gradients and generally converges quicker to a good minimum (as indicated by our training of circuits with 150 qubits on genomic data). Finally, as a by-product, the developed variance lower bound framework is applicable to a general class of non-linear losses, offering a broader toolset for analyzing warm-starts in quantum machine learning.

Evolve-based agent such as AlphaEvolve is one of the notable successes in using Large Language Models (LLMs) to build AI Scientists. These agents tackle open-ended scientific problems by iteratively improving and evolving programs, leveraging the prior knowledge and reasoning capabilities of LLMs. Despite the success, existing evolve-based agents lack targeted guidance for evolution and effective mechanisms for organizing and utilizing knowledge acquired from past evolutionary experience. Consequently, they suffer from decreasing evolution efficiency and exhibit oscillatory behavior when approaching known performance boundaries. To mitigate the gap, we develop CausalEvolve, equipped with a causal scratchpad that leverages LLMs to identify and reason about guiding factors for evolution. At the beginning, CausalEvolve first identifies outcome-level factors that offer complementary inspirations in improving the target objective. During the evolution, CausalEvolve also inspects surprise patterns during the evolution and abductive reasoning to hypothesize new factors, which in turn offer novel directions. Through comprehensive experiments, we show that CausalEvolve effectively improves the evolutionary efficiency and discovers better solutions in 4 challenging open-ended scientific tasks.

Dynamical Mean Field Theory (DMFT) provides an asymptotic description of the dynamics of macroscopic observables in certain disordered systems. Originally pioneered in the context of spin glasses by Sompolinsky and Zippelius (1982), it has since been used to derive asymptotic dynamical equations for a wide range of models in physics, high-dimensional statistics and machine learning. One of the main tools used by physicists to obtain these equations is the dynamical cavity method, which has remained largely non-rigorous. In contrast, existing mathematical formalizations have relied on alternative approaches, including Gaussian conditioning, large deviations over paths, or Fourier analysis. In this work, we formalize the dynamical cavity method and use it to give a new proof of the DMFT equations for General First Order Methods, a broad class of dynamics encompassing algorithms such as Gradient Descent and Approximate Message Passing.

Objective: This study investigates the perceived value and critique of ISTQB certifications, the most widely recognized testing qualifications worldwide. While the certifications aim to standardize the software testing body of knowledge, debates persist about their practical relevance and impact. Our objective was to systematically capture practitioner perspectives and assess the precision of endorsements and fairness of criticisms through expert review. Method: We conducted an AI-assisted Multivocal Literature Review (MLR), combining academic and grey literature to synthesize practitioner endorsements (RQ1) and criticisms (RQ2). ChatGPT's deep research capability was employed under continuous human oversight, with QA strategies ensuring transparency and reliability. As another analysis, we asked a panel of four independent experts to evaluate the precision of endorsements and fairness of criticisms. Results: Practitioner endorsements emphasized career benefits, improved communication, and a shared vocabulary as the main values of ISTQB certifications. Criticisms focused on excessive theoretical content, limited relevance in agile and automation-intensive contexts, and weak support for real testing skills. Expert review confirmed that while many endorsements were precise, several criticisms reflected broader tensions in the discipline, including contrasting schools of thought in testing practice. Conclusions: ISTQB certifications provide recognizable career and communication value but remain contested in terms of practical utility. By triangulating practitioner voices with expert validation, this study delivers an evidence-based reflection on the strengths and weaknesses of ISTQB in shaping the software testing body of knowledge. The AI-assisted MLR also demonstrates how GenAI tools can support systematic evidence synthesis when coupled with rigorous human oversight.

Probabilistic language generators are theoretically modeled as discrete stochastic processes, yet standard decoding strategies (Top-k, Top-p) impose static truncation rules that fail to accommodate the dynamic information density of natural language. This misalignment often forces a suboptimal trade-off: static bounds are either too restrictive for high-entropy creative generation or too permissive for low-entropy logical reasoning. In this work, we formalize the generation process as a trajectory through a relative probability manifold. We introduce Top-b (Adaptive Relative Band Sampling), a decoding strategy that regulates the candidate set via a dynamic bandwidth coefficient coupled strictly to the instantaneous Shannon entropy of the model's distribution. We provide a theoretical framework demonstrating that Top-b acts as a variance-minimizing operator on the tail distribution. Empirical validation on GPQA and GSM8K benchmarks indicates that Top-b significantly reduces generation entropy and inter-decoding variance while maintaining competitive reasoning accuracy, effectively approximating a self-regulating control system for autoregressive generation.

The development of robust language models for low-resource languages is frequently bottlenecked by the scarcity of high-quality, coherent, and domain-appropriate training corpora. In this paper, we introduce the Multilingual TinyStories dataset, a large-scale, synthetically generated collection of children's stories encompassing 17 Indian languages. Designed specifically for the training and evaluation of Small Language Models (SLMs), the corpus provides simple, narrative-driven text strictly localized to native scripts. We detail our hybrid curation pipeline, which leverages the Sarvam-M language model and a novel combinatorial prompt engineering framework for native generation, coupled with the Google Translate API for large-scale cross-lingual expansion. Through strict programmatic filtering, we compiled 132,942 stories and over 93.9 million tokens in our release, serving as a foundational resource for multilingual language modeling and transfer learning in the Indic linguistic sphere.

Ulcerative colitis (UC) is a chronic mucosal inflammatory condition that places patients at increased risk of colorectal cancer. Colonoscopic surveillance remains the gold standard for assessing disease activity, and reporting typically relies on standardised endoscopic scoring metrics. The most widely used is the Mayo Endoscopic Score (MES), with some centres also adopting the Ulcerative Colitis Endoscopic Index of Severity (UCEIS). Both are descriptive assessments of mucosal inflammation (MES: 0 to 3; UCEIS: 0 to 8), where higher values indicate more severe disease. However, computational methods for automatically predicting these scores remain limited, largely due to the lack of publicly available expert-annotated datasets and the absence of robust benchmarking. There is also a significant research gap in generating clinically meaningful descriptions of UC images, despite image captioning being a well-established computer vision task. Variability in endoscopic systems and procedural workflows across centres further highlights the need for multi-centre datasets to ensure algorithmic robustness and generalisability. In this work, we introduce a curated multi-centre, multi-resolution dataset that includes expert-validated MES and UCEIS labels, alongside detailed clinical descriptions. To our knowledge, this is the first comprehensive dataset that combines dual scoring metrics for classification tasks with expert-generated captions describing mucosal appearance and clinically accepted reasoning for image captioning. This resource opens new opportunities for developing clinically meaningful multimodal algorithms. In addition to the dataset, we also provide benchmarking using convolutional neural networks, vision transformers, hybrid models, and widely used multimodal vision-language captioning algorithms.

Recruiters and job seekers rely on search systems to navigate labor markets, making candidate matching engines critical for hiring outcomes. Most systems act as keyword filters, failing to handle skill synonyms and nonlinear careers, resulting in missed candidates and opaque match scores. We introduce JobMatchAI, a production-ready system integrating Transformer embeddings, skill knowledge graphs, and interpretable reranking. Our system optimizes utility across skill fit, experience, location, salary, and company preferences, providing factor-wise explanations through resume-driven search workflows. We release JobSearch-XS benchmark and a hybrid retrieval stack combining BM25, knowledge graph and semantic components to evaluate skill generalization. We assess system performance on JobSearch-XS across retrieval tasks, provide a demo video, a hosted website and installable package.

We present, to our knowledge, the first language-driven agent system capable of executing end-to-end collider phenomenology tasks, instantiated within a decoupled, domain-agnostic architecture for autonomous High-Energy Physics phenomenology. Guided only by natural-language prompts supplemented with standard physics notation, ColliderAgent carries out workflows from a theoretical Lagrangian to final phenomenological outputs without relying on package-specific code. In this framework, a hierarchical multi-agent reasoning layer is coupled to Magnus, a unified execution backend for phenomenological calculations and simulation toolchains. We validate the system on representative literature reproductions spanning leptoquark and axion-like-particle scenarios, higher-dimensional effective operators, parton-level and detector-level analyses, and large-scale parameter scans leading to exclusion limits. These results point to a route toward more automated, scalable, and reproducible research in collider physics, cosmology, and physics more broadly.

Task sequencing (TS) is one of the core open problems in Deep Learning, arising in a plethora of real-world domains, from robotic assembly lines to autonomous driving. Unfortunately, prior work has not convincingly demonstrated the generalization ability of meta-learned TS methods to solve new TS problems, given few initial demonstrations. In this paper, we demonstrate that deep neural networks can meta-learn over an infinite prior of synthetically generated TS problems and achieve a few-shot generalization. We meta-learn a transformer-based architecture over datasets of sequencing trajectories generated from a prior distribution that samples sequencing problems as paths in directed graphs. In a large-scale experiment, we provide ample empirical evidence that our meta-learned models discover optimal task sequences significantly quicker than non-meta-learned baselines.

While Large Vision-Language Models (LVLMs) demonstrate exceptional multi-modal capabilities, the quadratic computational cost of processing high-resolution visual tokens remains a critical bottleneck. Though recent token reduction strategies attempt to accelerate inference, such methods inadequately exploit attention values and fail to address token redundancy. More critically, they overlook the ``attention shift'' phenomenon inherent in LVLMs, which skews token attention scores. In this work, we propose ASAP, a novel training-free, KV-Cache-compatible pruning recipe that comprehensively addresses these limitations. First, we mitigate the attention shift by utilizing a dynamic bidirectional soft attention mask, ensuring the selection of genuinely informative tokens rather than naive attention-based selection. Second, we posit that high semantic redundancy within the token set degrades performance. We therefore introduce a weighted soft merging component that merges semantically similar tokens, preserving only the most feature-dense visual patches for subsequent layers. ASAP achieves virtually lossless compression of visual context, retaining 99.02% of the original LLaVA-NeXT-7B performance while aggressively slashing computational FLOPs by ~80%.

The departure of subject-matter experts from industrial organizations results in the irreversible loss of tacit knowledge that is rarely captured through conventional documentation practices. This paper proposes Expert Mind, an experimental system that leverages Retrieval-Augmented Generation (RAG), large language models (LLMs), and multimodal capture techniques to preserve, structure, and make queryable the deep expertise of organizational knowledge holders. Drawing on the specific context of the energy sector, where decades of operational experience risk being lost to an aging workforce, we describe the system architecture, processing pipeline, ethical framework, and evaluation methodology. The proposed system addresses the knowledge elicitation problem through structured interviews, think-aloud sessions, and text corpus ingestion, which are subsequently embedded into a vector store and queried through a conversational interface. Preliminary design considerations suggest Expert Mind can significantly reduce knowledge transfer latency and improve onboarding efficiency. Ethical dimensions including informed consent, intellectual property, and the right to erasure are addressed as first-class design constraints.

Reinforcement learning has proven its power on various occasions. However, its performance is not always guaranteed when system dynamics change. Instead, it largely relies on users' empirical experience. For reinforcement learning algorithms with an actor-critic structure, the critic neural network reflects the approximation and optimization process in the RL algorithm. Analyzing the performance of the critic neural network helps to understand the mechanism of the algorithm. To support systematic interpretation of such algorithms in dynamic control problems, this work proposes a critic match loss landscape visualization method for online reinforcement learning. The method constructs a loss landscape by projecting recorded critic parameter trajectories onto a low-dimensional linear subspace. The critic match loss is evaluated over the projected parameter grid using fixed reference state samples and temporal-difference targets. This yields a three-dimensional loss surface together with a two-dimensional optimization path that characterizes critic learning behavior. To extend analysis beyond visual inspection, quantitative landscape indices and a normalized system performance index are introduced, enabling structured comparison across different training outcomes. The approach is demonstrated using the Action-Dependent Heuristic Dynamic Programming algorithm on cart-pole and spacecraft attitude control tasks. Comparative analyses across projection methods and training stages reveal distinct landscape characteristics associated with stable convergence and unstable learning. The proposed framework enables both qualitative and quantitative interpretation of critic optimization behavior in online reinforcement learning.

Humans learn from catastrophic mistakes not through numerical penalties, but through qualitative suffering that reshapes who they are. Current AI safety approaches replicate none of this. Reward shaping captures magnitude, not meaning. Rule-based alignment constrains behaviour, but does not change it. We propose Emotional Cost Functions, a framework in which agents develop Qualitative Suffering States, rich narrative representations of irreversible consequences that persist forward and actively reshape character. Unlike numerical penalties, qualitative suffering states capture the meaning of what was lost, the specific void it creates, and how it changes the agent's relationship to similar future situations. Our four-component architecture - Consequence Processor, Character State, Anticipatory Scan, and Story Update is grounded in one principle. Actions cannot be undone and agents must live with what they have caused. Anticipatory dread operates through two pathways. Experiential dread arises from the agent's own lived consequences. Pre-experiential dread is acquired without direct experience, through training or inter-agent transmission. Together they mirror how human wisdom accumulates across experience and culture. Ten experiments across financial trading, crisis support, and content moderation show that qualitative suffering produces specific wisdom rather than generalised paralysis. Agents correctly engage with moderate opportunities at 90-100% while numerical baselines over-refuse at 90%. Architecture ablation confirms the mechanism is necessary. The full system generates ten personal grounding phrases per probe vs. zero for a vanilla LLM. Statistical validation (N=10) confirms reproducibility at 80-100% consistency.

The biochemical processes underlying DNA data storage, including synthesis, amplification, and sequencing, are inherently noisy. Consequently, base-level insertion, deletion, and substitution (IDS) errors, as well as sequence-level dropouts, occur and pose major challenges for reliable data retrieval. Here we introduce DNA-MGC+, a DNA storage codec designed to enable reliable and resource-efficient data retrieval under diverse operating conditions. We evaluate DNA-MGC+ across a wide range of in silico and in vitro settings, including experiments with both Illumina and Nanopore sequencing, and show that it consistently outperforms existing codecs. In particular, DNA-MGC+ achieves simultaneous gains in sequencing depth requirements, read cost, decoding time, storage density, and error-correction capability under explicit reliability constraints. Notable results include reliable decoding under IDS error rates of up to 24% in synthetic scenarios, and reliable retrieval at sequencing depths below 3x with read costs below 3.5 bits/nt under electrochemical synthesis for both Illumina and Nanopore sequencing.

The intentional creation and spread of disinformation poses a significant threat to public discourse. However, existing English datasets and research rarely address the intentionality behind the disinformation. This work presents MALINT, the first human-annotated English corpus developed in collaboration with expert fact-checkers to capture disinformation and its malicious intent. We utilize our novel corpus to benchmark 12 language models, including small language models (SLMs) such as BERT and large language models (LLMs) like Llama 3.3, on binary and multilabel intent classification tasks. Moreover, inspired by inoculation theory from psychology and communication studies, we investigate whether incorporating knowledge of malicious intent can improve disinformation detection. To this end, we propose intent-based inoculation, an intent-augmented reasoning for LLMs that integrates intent analysis to mitigate the persuasive impact of disinformation. Analysis on six disinformation datasets, five LLMs, and seven languages shows that intent-augmented reasoning improves zero-shot disinformation detection. To support research in intent-aware disinformation detection, we release the MALINT dataset with annotations from each annotation step.

Vision-Language-Action (VLA) models have shown promising capabilities for embodied intelligence, but most existing approaches rely on text-based chain-of-thought reasoning where visual inputs are treated as static context. This limits the ability of the model to actively revisit the environment and resolve ambiguities during long-horizon tasks. We propose VLA-Thinker, a thinking-with-image reasoning framework that models perception as a dynamically invocable reasoning action. To train such a system, we introduce a two-stage training pipeline consisting of (1) an SFT cold-start phase with curated visual Chain-of-Thought data to activate structured reasoning and tool-use behaviors, and (2) GRPO-based reinforcement learning to align complete reasoning-action trajectories with task-level success. Extensive experiments on LIBERO and RoboTwin 2.0 benchmarks demonstrate that VLA-Thinker significantly improves manipulation performance, achieving 97.5% success rate on LIBERO and strong gains across long-horizon robotic tasks. Project and Codes: https://cywang735.github.io/VLA-Thinker/ .

Large language models (LLMs) suffer from proactive interference (PI): outdated information in the context window disrupts retrieval of current values. This interference degrades retrieval accuracy log-linearly as stale associations accumulate, a bottleneck that persists regardless of context length and resists prompt-engineering mitigations. Biological brains resolve an analogous challenge through sleep-dependent memory consolidation: synaptic downscaling, selective replay, and targeted forgetting. We propose SleepGate, a biologically inspired framework that augments transformer-based LLMs with a learned sleep cycle over the key-value (KV) cache. SleepGate introduces three mechanisms: (1) a conflict-aware temporal tagger detecting when new entries supersede old ones; (2) a lightweight forgetting gate trained to selectively evict or compress stale cache entries; and (3) a consolidation module that merges surviving entries into compact summaries. These components activate periodically during inference in sleep micro-cycles, governed by an adaptive entropy-based trigger. We formalize a dual-phase training objective jointly optimizing language modeling during the wake phase and post-consolidation retrieval during the sleep phase. Theoretical analysis shows SleepGate reduces the interference horizon from O(n) to O(log n). In experiments with a small-scale transformer (4 layers, 793K parameters), SleepGate achieves 99.5% retrieval accuracy at PI depth 5 and 97.0% at depth 10, while all five baselines -- full KV cache, sliding window, H2O, StreamingLLM, and decay-only ablation -- remain below 18%. Our framework offers an architecture-level solution that prompt engineering cannot address.

Neural-network wave functions in Variational Monte Carlo (VMC) have achieved great success in accurately representing both ground and excited states. However, achieving sufficient numerical accuracy in state overlaps requires increasing the number of Monte Carlo samples, and consequently the computational cost, with the number of states. We present a nearly constant sample-size approach, Multi-State Importance Sampling (MSIS), that leverages samples from all states to estimate pairwise overlap. To efficiently evaluate all states for all samples, we introduce Excited Pfaffians. Inspired by Hartree-Fock, this architecture represents many states within a single neural network. Excited Pfaffians also serve as generalized wave functions, allowing a single model to represent multi-state potential energy surfaces. On the carbon dimer, we match the $O(N_s^4)$-scaling natural excited states while training $>200\times$ faster and modeling 50\% more states. Our favorable scaling enables us to be the first to use neural networks to find all distinct energy levels of the beryllium atom. Finally, we demonstrate that a single wave function can represent excited states across various molecules.

We present the first uniform-in-time high-probability bound for SGD under the PL condition, where the gradient noise contains both Markovian and martingale difference components. This significantly broadens the scope of finite-time guarantees, as the PL condition arises in many machine learning and deep learning models while Markovian noise naturally arises in decentralized optimization and online system identification problems. We further allow the magnitude of noise to grow with the function value, enabling the analysis of many practical sampling strategies. In addition to the high-probability guarantee, we establish a matching $1/k$ decay rate for the expected suboptimality. Our proof technique relies on the Poisson equation to handle the Markovian noise and a probabilistic induction argument to address the lack of almost-sure bounds on the objective. Finally, we demonstrate the applicability of our framework by analyzing three practical optimization problems: token-based decentralized linear regression, supervised learning with subsampling for privacy amplification, and online system identification.

Optimizing the noise samples of diffusion and flow models is an increasingly popular approach to align these models to target rewards at inference time. However, we observe that these approaches are usually restricted to differentiable or cheap reward models, the formulation of the underlying pretrained generative model, or are memory/compute inefficient. We instead propose a simple trust-region based search algorithm (TRS) which treats the pre-trained generative and reward models as a black-box and only optimizes the source noise. Our approach achieves a good balance between global exploration and local exploitation, and is versatile and easily adaptable to various generative settings and reward models with minimal hyperparameter tuning. We evaluate TRS across text-to-image, molecule and protein design tasks, and obtain significantly improved output samples over the base generative models and other inference-time alignment approaches which optimize the source noise sample, or even the entire reverse-time sampling noise trajectories in the case of diffusion models. Our source code is publicly available.

Large Language Models excel in high-resource programming languages but struggle with low-resource ones. Existing research related to low-resource programming languages primarily focuses on Domain-Specific Languages (DSLs), leaving general-purpose languages that suffer from data scarcity underexplored. To address this gap, we introduce CangjieBench, a contamination-free benchmark for Cangjie, a representative low-resource general-purpose language. The benchmark comprises 248 high-quality samples manually translated from HumanEval and ClassEval, covering both Text-to-Code and Code-to-Code tasks. We conduct a systematic evaluation of diverse LLMs under four settings: Direct Generation, Syntax-Constrained Generation, Retrieval-Augmented Generation (RAG), and Agent. Experiments reveal that Direct Generation performs poorly, whereas Syntax-Constrained Generation offers the best trade-off between accuracy and computational cost. Agent achieve state-of-the-art accuracy but incur high token consumption. Furthermore, we observe that Code-to-Code translation often underperforms Text-to-Code generation, suggesting a negative transfer phenomenon where models overfit to the source language patterns. We hope that our work will offer valuable insights into LLM generalization to unseen and low-resource programming languages. Our code and data are available at https://github.com/cjhCoder7/CangjieBench.

Accurate 3D anatomical segmentation is essential for clinical diagnosis and surgical planning. However, automated models frequently generate suboptimal shape predictions due to factors such as limited and imbalanced training data, inadequate labeling quality, and distribution shifts between training and deployment settings. A natural solution is to iteratively refine the predicted shape based on the radiologists' verbal instructions. However, this is hindered by the scarcity of paired data that explicitly links erroneous shapes to corresponding corrective instructions. As an initial step toward addressing this limitation, we introduce CoWTalk, a benchmark comprising 3D arterial anatomies with controllable synthesized anatomical errors and their corresponding repairing instructions. Building on this benchmark, we further propose an iterative refinement model that represents 3D shapes as vector sets and interacts with textual instructions to progressively update the target shape. Experimental results demonstrate that our method achieves significant improvements over corrupted inputs and competitive baselines, highlighting the feasibility of language-driven clinician-in-the-loop refinement for 3D medical shapes modeling.

Tensions between AI Safety (AIS) and AI Ethics (AIE) have increasingly surfaced in AI governance and public debates about AI, leading to what we term the "responsible AI divides". We introduce a model that categorizes four modes of engagement with the tensions: radical confrontation, disengagement, compartmentalized coexistence, and critical bridging. We then investigate how critical bridging, with a particular focus on bridging problems, offers one of the most viable constructive paths for advancing responsible AI. Using computational tools to analyze a curated dataset of 3,550 papers, we map the research landscapes of AIE and AIS to identify both distinct and overlapping problems. Our findings point to both thematic divides and overlaps. For example, we find that AIE has long grappled with overcoming injustice and tangible AI harms, whereas AIS has primarily embodied an anticipatory approach focused on the mitigation of risks from AI capabilities. At the same time, we find significant overlap in core research concerns across both AIE and AIS around transparency, reproducibility, and inadequate governance mechanisms. As AIE and AIS continue to evolve, we recommend focusing on bridging problems as a constructive path forward for enhancing collaborative AI governance. We offer a series of recommendations to integrate shared considerations into a collaborative approach to responsible AI. Alongside our proposal, we highlight its limitations and explore open problems for future research. All data including the fully annotated dataset of papers with code to reproduce our figures can be found at: https://github.com/gyevnarb/ai-safety-ethics.

The paper demonstrate that simple adjustments of the fine-tuning recipes of multimodal large language models (MLLM) are sufficient to mitigate catastrophic forgetting. On visual question answering, we design a 2x2 experimental framework to assess model performance across in-distribution and out-of-distribution image and text inputs. Our results show that appropriate regularization, such as constraining the number of trainable parameters or adopting a low learning rate, effectively prevents forgetting when dealing with out-of-distribution images. However, we uncover a distinct form of forgetting in settings with in-distribution images and out-of-distribution text. We attribute this forgetting as task-specific overfitting and address this issue by introducing a data-hybrid training strategy that combines datasets and tasks. Finally, we demonstrate that this approach naturally extends to continual learning, outperforming existing methods with complex auxiliary mechanisms. In general, our findings challenge the prevailing assumptions by highlighting the inherent robustness of MLLMs and providing practical guidelines for adapting them while preserving their general capabilities.

Predicting the stress-strain behaviors of additively manufactured materials is crucial for part qualification in additive manufacturing (AM). Conventional physics-based constitutive models often oversimplify material properties, while data-driven machine learning (ML) models often lack physical consistency and interpretability. To address these issues, we propose a physics-informed machine learning (PIML) framework to improve the predictive performance and physical consistency for predicting the stress-strain curves of additively manufactured polymers and metals. A polynomial regression model is used to predict the yield point from AM process parameters, then stress-strain curves are segmented into elastic and plastic regions. Two long short-term memory (LSTM) models are trained to predict two regions separately. For the elastic region, Hooke's law is embedded into the LSTM model for both polymer and metal. For the plastic region, Voce hardening law and Hollomon's law are embedded into the LSTM model for polymer and metal, respectively. The loss-based and activation-based PIML architectures are developed by embedding the physical laws into the loss and activation functions, respectively. The performance of the two PIML architectures are compared with two LSTM-based ML models, three additional ML models, and a physics-based constitutive model. These models are built on experimental data collected from two additively manufactured polymers (i.e., Nylon and carbon fiber-acrylonitrile butadiene styrene) and two additively manufactured metals (i.e., AlSi10Mg and Ti6Al4V). Experimental results demonstrate that two PIML architectures consistently outperform the other models. The segmental predictive model with activation-based PIML architecture achieves the lowest MAPE of 10.46+/-0.81% and the highest R^2 of 0.82+/-0.05 arocss four datasets.

Training large language models for complex reasoning is bottlenecked by the scarcity of verifiable, high-quality data. In domains like physics, standard text augmentation often introduces hallucinations, while static benchmarks lack the reasoning traces required for fine-tuning. We introduce the Infinite Problem Generator (IPG), an agentic framework that synthesizes physics problems with guaranteed solvability through a Formula-as-Code paradigm. Unlike probabilistic text generation, IPG constructs solutions as executable Python programs, enforcing strict mathematical consistency. As a proof-of-concept, we release ClassicalMechanicsV1, a high-fidelity corpus of 1,335 classical mechanics problems expanded from 165 expert seeds. The corpus demonstrates high structural diversity, spanning 102 unique physical formulas with an average complexity of 3.05 formulas per problem. Furthermore, we identify a Complexity Blueprint, demonstrating a strong linear correlation ($R^2 \approx 0.95$) between formula count and verification code length. This relationship establishes code complexity as a precise, proxy-free metric for problem difficulty, enabling controllable curriculum generation. We release the full IPG pipeline, the ClassicalMechanicsV1 dataset, and our evaluation report to support reproducible research in reasoning-intensive domains.

Traditional machine learning assumes a stationary data distribution, yet many real-world applications operate on nonstationary streams in which the underlying concept evolves over time. This problem can also be viewed as task-free continual learning under concept drift, where a model must adapt sequentially without explicit task identities or task boundaries. In such settings, effective learning requires both rapid adaptation to new data and forgetting of outdated information. A common solution is based on a sliding window, but this approach is often computationally demanding because the model must be repeatedly retrained from scratch on the most recent data. We propose a different perspective based on machine unlearning. Instead of rebuilding the model each time the active window changes, we remove the influence of outdated samples using unlearning and then update the model with newly observed data. This enables efficient, targeted forgetting while preserving adaptation to evolving distributions. To the best of our knowledge, this is the first work to connect machine unlearning with concept drift mitigation for task-free continual learning. Empirical results on image stream classification across multiple drift scenarios demonstrate that the proposed approach offers a competitive and computationally efficient alternative to standard sliding-window retraining. Our implementation can be found at \hrehttps://anonymous.4open.science/r/MUNDataStream-60F3}{https://anonymous.4open.science/r/MUNDataStream-60F3}.

Parametric system identification methods estimate the parameters of explicitly defined physical systems from data. Yet, they remain constrained by the need to provide an explicit function space, typically through a predefined library of candidate functions chosen via available domain knowledge. In contrast, deep learning can demonstrably model systems of broad complexity with high fidelity, but black-box function approximation typically fails to yield explicit descriptive or disentangled representations revealing the structure of a system. We develop a novel identifiability theorem, leveraging causal representation learning, to uncover disentangled representations of system parameters without structural assumptions. We derive a graphical criterion specifying when system parameters can be uniquely disentangled from raw trajectory data, up to permutation and diffeomorphism. Crucially, our analysis demonstrates that global causal structures provide a lower bound on the disentanglement guarantees achievable when considering local state-dependent causal structures. We instantiate system parameter identification as a variational inference problem, leveraging a sparsity-regularised transformer to uncover state-dependent causal structures. We empirically validate our approach across four synthetic domains, demonstrating its ability to recover highly disentangled representations that baselines fail to recover. Corroborating our theoretical analysis, our results confirm that enforcing local causal structure is often necessary for full identifiability.

In this paper, we analyze the two time-scale stochastic approximation (TTSSA) algorithm introduced in Borkar (1997) using a martingale approach. This approach leads to simple sufficient conditions for the iterations to be bounded almost surely, as well as estimates on the rate of convergence of the mean-squared error of the TTSSA algorithm to zero. Our theory is applicable to nonlinear equations, in contrast to many papers in the TTSSA literature which assume that the equations are linear. The convergence of TTSSA is proved in the "almost sure" sense, in contrast to earlier papers on TTSSA that establish convergence in distribution, convergence in the mean, and the like. Moreover, in this paper we establish different rates of convergence for the fast and the slow subsystems, perhaps for the first time. Finally, all of the above results to continue to hold in the case where the two measurement errors have nonzero conditional mean, and/or have conditional variances that grow without bound as the iterations proceed. This is in contrast to previous papers which assumed that the errors form a martingale difference sequence with uniformly bounded conditional variance. It is shown that when the measurement errors have zero conditional mean and the conditional variance remains bounded, the mean-squared error of the iterations converges to zero at a rate of $o(t^{-η})$ for all $η\in (0,1)$. This improves upon the rate of $O(t^{-2/3})$ proved in Doan (2023) (which is the best bound available to date). Our bound is virtually the same as the rate of $O(t^{-1})$ proved in Doan (2024), but for a Polyak-Ruppert averaged version of TTSSA, and not directly. Rates of convergence are also established for the case where the errors have nonzero conditional mean and/or unbounded conditional variance.

This study presents a geometric deep learning framework for predicting cold spray particle impact responses using finite element simulation data. A parametric dataset was generated through automated Abaqus simulations spanning a systematic range of particle velocity, particle temperature, and friction coefficient, yielding five output targets including maximum equivalent plastic strain, average contact plastic strain, maximum temperature, maximum von Mises stress, and deformation ratio. Four novel algorithms i.e. a GraphSAGE-style inductive graph neural network, a Chebyshev spectral graph convolution network, a topological data analysis augmented multilayer perceptron, and a geometric attention network were implemented and evaluated. Each input sample was treated as a node in a k-nearest-neighbour feature-space graph, enabling the models to exploit spatial similarity between process conditions during training. Three-dimensional feature space visualisations and two-dimensional contour projections confirmed the highly non-linear and velocity-dominated nature of the input-output relationships. Quantitative evaluation demonstrated that GraphSAGE and GAT consistently achieved R-square values exceeding 0.93 across most targets, with GAT attaining peak performance of R-square equal to 0.97 for maximum plastic strain. ChebSpectral and TDA-MLP performed considerably worse, yielding negative R-square values for several targets. These findings establish spatial graph-based neighbourhood aggregation as a robust and physically interpretable surrogate modelling strategy for cold spray process optimisation.

Sketch techniques have been extensively studied in recent years and are especially well-suited to data streaming scenarios, where the sketch summary is updated quickly and compactly. However, it is challenging to recover the current state from these summaries in a way that is accurate, fast, and real. In this paper, we seek a solution that reconciles this tension, aiming for near-perfect recovery with lightweight computational procedures. Focusing on linear sketching problems of the form $\boldsymbolΦf \rightarrow f$, our study proceeds in three stages. First, we dissect existing techniques and show the root cause of the sketching dilemma: an orthogonal information loss. Second, we examine how generative priors can be leveraged to bridge the information gap. Third, we propose FLORE, a novel generative sketching framework that embraces these analyses to achieve the best of all worlds. More importantly, FLORE can be trained without access to ground-truth data. Comprehensive evaluations demonstrate FLORE's ability to provide high-quality recovery, and support summary with low computing overhead, outperforming previous methods by up to 1000 times in error reduction and 100 times in processing speed compared to learning-based solutions.

Great scientists have strong judgement and foresight, closely tied to what we call scientific taste. Here, we use the term to refer to the capacity to judge and propose research ideas with high potential impact. However, most relative research focuses on improving an AI scientist's executive capability, while enhancing an AI's scientific taste remains underexplored. In this work, we propose Reinforcement Learning from Community Feedback (RLCF), a training paradigm that uses large-scale community signals as supervision, and formulate scientific taste learning as a preference modeling and alignment problem. For preference modeling, we train Scientific Judge on 700K field- and time-matched pairs of high- vs. low-citation papers to judge ideas. For preference alignment, using Scientific Judge as a reward model, we train a policy model, Scientific Thinker, to propose research ideas with high potential impact. Experiments show Scientific Judge outperforms SOTA LLMs (e.g., GPT-5.2, Gemini 3 Pro) and generalizes to future-year test, unseen fields, and peer-review preference. Furthermore, Scientific Thinker proposes research ideas with higher potential impact than baselines. Our findings show that AI can learn scientific taste, marking a key step toward reaching human-level AI scientists.

Reinforcement learning (RL) has achieved strong performance in robotic control; however, state-of-the-art policy learning methods, such as actor-critic methods, still suffer from high sample complexity and often produce physically inconsistent actions. This limitation stems from neural policies implicitly rediscovering complex physics from data alone, despite accurate dynamics models being readily available in simulators. In this paper, we introduce a novel physics-informed RL framework, called PIPER, that seamlessly integrates physical constraints directly into neural policy optimization with analytical soft physics constraints. At the core of our method is the integration of a differentiable Lagrangian residual as a regularization term within the actor's objective. This residual, extracted from a robot's simulator description, subtly biases policy updates towards dynamically consistent solutions. Crucially, this physics integration is realized through an additional loss term during policy optimization, requiring no alterations to existing simulators or core RL algorithms. Extensive experiments demonstrate that our method significantly improves learning efficiency, stability, and control accuracy, establishing a new paradigm for efficient and physically consistent robotic control.

While Large Language Models (LLMs) have evolved into tool-using agents, they remain brittle in long-horizon interactions. Unlike mathematical reasoning where errors are often rectifiable via backtracking, tool-use failures frequently induce irreversible side effects, making accurate step-level verification critical. However, existing process-level benchmarks are predominantly confined to closed-world mathematical domains, failing to capture the dynamic and open-ended nature of tool execution. To bridge this gap, we introduce AgentProcessBench, the first benchmark dedicated to evaluating step-level effectiveness in realistic, tool-augmented trajectories. The benchmark comprises 1,000 diverse trajectories and 8,509 human-labeled step annotations with 89.1% inter-annotator agreement. It features a ternary labeling scheme to capture exploration and an error propagation rule to reduce labeling ambiguity. Extensive experiments reveal key insights: (1) weaker policy models exhibit inflated ratios of correct steps due to early termination; (2) distinguishing neutral and erroneous actions remains a significant challenge for current models; and (3) process-derived signals provide complementary value to outcome supervision, significantly enhancing test-time scaling. We hope AgentProcessBench can foster future research in reward models and pave the way toward general agents. The code and data are available at https://github.com/RUCBM/AgentProcessBench.

Adapting Large Language Models (LLMs) to high-stakes vertical domains like insurance presents a significant challenge: scenarios demand strict adherence to complex regulations and business logic with zero tolerance for hallucinations. Existing approaches often suffer from a Competency Trade-off - sacrificing general intelligence for domain expertise - or rely heavily on RAG without intrinsic reasoning. To bridge this gap, we present INS-S1, an insurance-specific LLM family trained via a novel end-to-end alignment paradigm. Our approach features two methodological innovations: (1) A Verifiable Data Synthesis System that constructs hierarchical datasets for actuarial reasoning and compliance; and (2) A Progressive SFT-RL Curriculum Framework that integrates dynamic data annealing with a synergistic mix of Verified Reasoning (RLVR) and AI Feedback (RLAIF). By optimizing data ratios and reward signals, this framework enforces domain constraints while preventing catastrophic forgetting. Additionally, we release INSEva, the most comprehensive insurance benchmark to date (39k+ samples). Extensive experiments show that INS-S1 achieves SOTA performance on domain tasks, significantly outperforming DeepSeek-R1 and Gemini-2.5-Pro. Crucially, it maintains top-tier general capabilities and achieves a record-low 0.6% hallucination rate (HHEM). Our results demonstrate that rigorous domain specialization can be achieved without compromising general intelligence.

Honey bee colony losses threaten global pollination services, yet current monitoring systems treat each hive as an isolated unit, ignoring the spatial pathways through which diseases spread across apiaries. This paper introduces the Spatio-Temporal Apiary Graph Convolutional Network (STAG-CN), a graph neural network that models inter-hive relationships for disease onset prediction. STAG-CN operates on a dual adjacency graph combining physical co-location and climatic sensor correlation among hive sessions, and processes multivariate IoT sensor streams through a temporal--spatial--temporal sandwich architecture built on causal dilated convolutions and Chebyshev spectral graph convolutions. Evaluated on the Korean AI Hub apiculture dataset (dataset \#71488) with expanding-window temporal cross-validation, STAG-CN achieves an F1 score of 0.607 at a three-day forecast horizon. An ablation study reveals that the climatic adjacency matrix alone matches full-model performance (F1\,=\,0.607), while the physical adjacency alone yields F1\,=\,0.274, indicating that shared environmental response patterns carry stronger predictive signal than spatial proximity for disease onset. These results establish a proof-of-concept for graph-based biosecurity monitoring in precision apiculture, demonstrating that inter-hive sensor correlations encode disease-relevant information invisible to single-hive approaches.

Artificial agents that support human group interactions hold great promise, especially in sensitive contexts such as well-being promotion and therapeutic interventions. However, current systems struggle to mediate group interactions involving people who are not neurotypical. This limitation arises because most AI detection models (e.g., for turn-taking) are trained on data from neurotypical populations. This work takes a step toward inclusive AI by addressing the challenge of eye contact detection, a core component of non-verbal communication, with and for people with Intellectual and Developmental Disabilities. First, we introduce a new dataset, Multi-party Interaction with Intellectual and Developmental Disabilities (MIDD), capturing atypical gaze and engagement patterns. Second, we present the results of a comparative analysis with neurotypical datasets, highlighting differences in class imbalance, speaking activity, gaze distribution, and interaction dynamics. Then, we evaluate classifiers ranging from SVMs to FSFNet, showing that fine-tuning on MIDD improves performance, though notable limitations remain. Finally, we present the insights gathered through a focus group with six therapists to interpret our quantitative findings and understand the practical implications of atypical gaze and engagement patterns. Based on these results, we discuss data-driven strategies and emphasize the importance of feature choice for building more inclusive human-centered tools.

Fact-seeking question answering with large language models (LLMs) remains unreliable when answers depend on up-to-date or conflicting information. Although retrieval-augmented and tool-using LLMs reduce hallucinations, they often rely on implicit planning, leading to inefficient tool usage. We propose a modular framework that explicitly separates planning from factual retrieval and answer synthesis. A lightweight student planner is trained via a teacher-student framework to generate structured decompositions consisting of abstract reasoning steps and searchable fact requests. The supervision signals contain only planning traces and fact requests, without providing factual answers or retrieved evidence. At inference, the planner produces plans, while prompt-engineered modules perform retrieval and response synthesis. We evaluate the proposed framework on SEAL-0, an extremely challenging benchmark for search-augmented LLMs. Results show that supervised planning improves both accuracy and latency compared to monolithic reasoning models and prompt-based tool-augmented frameworks, demonstrating that explicitly learned planning structures are essential for reliable fact-seeking LLMs.

Persian poses unique audio understanding challenges through its classical poetry, traditional music, and pervasive code-switching - none captured by existing benchmarks. We introduce PARSA-Bench (Persian Audio Reasoning and Speech Assessment Benchmark), the first benchmark for evaluating large audio-language models on Persian language and culture, comprising 16 tasks and over 8,000 samples across speech understanding, paralinguistic analysis, and cultural audio understanding. Ten tasks are newly introduced, including poetry meter and style detection, traditional Persian music understanding, and code-switching detection. Text-only baselines consistently outperform audio counterparts, suggesting models may not leverage audio-specific information beyond what transcription alone provides. Culturally-grounded tasks expose a qualitatively distinct failure mode: all models perform near random chance on vazn detection regardless of scale, suggesting prosodic perception remains beyond the reach of current models. The dataset is publicly available at https://huggingface.co/datasets/MohammadJRanjbar/PARSA-Bench

Whether parameterized quantum circuits (PQCs) can be systematically constructed to be both trainable and expressive remains an open question. Highly expressive PQCs often exhibit barren plateaus, while several trainable alternatives admit efficient classical simulation. We address this question by deriving a finite-sample, dimension-independent concentration bound for estimating the variance of a PQC cost function, yielding explicit trainability guarantees. Across commonly used ansätze, we observe an anticorrelation between trainability and expressibility, consistent with theoretical insights. Building on this observation, we propose a property-based ansatz-search framework for identifying circuits that combine trainability and expressibility. We demonstrate its practical viability on a real quantum computer and apply it to variational quantum algorithms. We identify quantum neural network ansätze with improved effective dimension using over $6 \times$ fewer parameters, and for VQE on $\mathrm{H}_2$ we achieve UCCSD-like accuracy at substantially reduced circuit complexity.

Multimodal Large Language Model (MLLM)-based Graphical User Interface (GUI) agents develop rapidly, with visual grounding that maps natural language instructions to target UI elements serving as the core capability. Existing GUI agents typically fine-tune MLLM on massive datasets to handle challenges in understanding instructions and UI interfaces, which not only incurs high data annotation costs but also makes performance dependent on data quality and distribution. To avoid such cumbersome yet ineffective training, we notice that complex UI interfaces can be decomposed into basic visual elements directly understandable by common MLLMs. Consequently, we propose ZoomUI that leverages inference scaling to guide common MLLMs in progressively anchor instruction elements to increasingly detailed interface elements. Specifically, ZoomUI first optimizes the latent thinking to transform original instruction into element visual features description, and subsequently leverages internal attention to iteratively zoom in target element interface region. Evaluations on extensive benchmarks demonstrate that ZoomUI reaches or even surpasses SOTA baselines.

Parallel Byzantine Fault Tolerant (BFT) protocols based on committee-based sharding improve scalability but weaken safety since smaller node groups are responsible for consensus. Recent approaches integrate trusted execution environments (TEEs) into parallel BFT frameworks to enhance safety. While the scalability and safety issues are addressed by trusted parallel BFT, existing committee configuration methods often rely on randomized assignment, which can degrade performance. This paper proposes a committee configuration optimization (CCO) model based on mixed integer programming to improve transaction performance for trusted parallel BFT. The model considers communication delays and node failure rates to determine an optimal committee configuration that minimizes transaction latency under both normal operations and scenarios of trusted hardware failures. We integrate CCO into a trusted parallel BFT protocol and evaluate the performance on Microsoft virtual machines. Experimental results demonstrate 15% and 21% improved transaction throughput under normal operations and fallback process, respectively, highlighting the benefits of optimization-driven committee configuration in trusted parallel BFT systems.

This study examines phonetic texture in Persian poetry as a literary-historical phenomenon rather than a by-product of meter or a feature used only for classification. The analysis draws on a large corpus of 1,116,306 mesras from 31,988 poems written by 83 poets, restricted to five major classical meters to enable controlled comparison. Each line is converted into a grapheme-to-phoneme representation and analyzed using six phonetic metrics: hardness, sonority, sibilance, vowel ratio, phoneme entropy, and consonant-cluster ratio. Statistical models estimate poet-level differences while controlling for meter, poetic form, and line length. The results show that although meter and form explain a substantial portion of phonetic variation, they do not eliminate systematic differences between poets. Persian poetic sound therefore appears as conditioned variation within shared prosodic structures rather than as either purely individual style or simple metrical residue. A multidimensional stylistic map reveals several recurrent phonetic profiles, including high-sonority lyric styles, hardness-driven rhetorical or epic styles, sibilant mystical contours, and high-entropy complex textures. Historical analysis indicates that phonetic distributions shift across centuries, reflecting changes in genre prominence, literary institutions, and performance contexts rather than abrupt stylistic breaks. The study establishes a corpus-scale framework for phonetic analysis in Persian poetry and demonstrates how computational phonetics can contribute to literary-historical interpretation while remaining attentive to the formal structures that shape Persian verse.

Blind source separation (BSS) seeks to recover latent source signals from observed mixtures. Variational autoencoders (VAEs) offer a natural perspective for this problem: the latent variables can be interpreted as source components, the encoder can be viewed as a demixing mapping from observations to sources, and the decoder can be regarded as a remixing process from inferred sources back to observations. In this work, we propose AR-Flow VAE, a novel VAE-based framework for BSS in which each latent source is endowed with a parameter-adaptive autoregressive flow prior. This prior significantly enhances the flexibility of latent source modeling, enabling the framework to capture complex non-Gaussian behaviors and structured dependencies, such as temporal correlations, that are difficult to represent with conventional priors. In addition, the structured prior design assigns distinct priors to different latent dimensions, thereby encouraging the latent components to separate into different source signals under heterogeneous prior constraints. Experimental results validate the effectiveness of the proposed architecture for blind source separation. More importantly, this work provides a foundation for future investigations into the identifiability and interpretability of AR-Flow VAE.

Large Language Models (LLMs) have demonstrated remarkable capabilities in narrative generation. However, they often produce structurally homogenized stories, frequently following repetitive arrangements and combinations of plot events along with stereotypical resolutions. In this paper, we propose a novel theoretical framework for analysis by incorporating Proppian narratology and narrative functions. This framework is used to analyze the composition of narrative texts generated by LLMs to uncover their underlying narrative logic. Taking Chinese web literature as our research focus, we extend Propp's narrative theory, defining 34 narrative functions suited to modern web narrative structures. We further construct a human-annotated corpus to support the analysis of narrative structures within LLM-generated text. Experiments reveal that the primary reasons for the singular narrative logic and severe homogenization in generated texts are that current LLMs are unable to correctly comprehend the meanings of narrative functions and instead adhere to rigid narrative generation paradigms.

Modern recommendation systems rank candidates by aggregating multiple behavioral signals through a value model. However, many commonly used signals are inherently affected by heterogeneous biases. For example, watch time naturally favors long-form content, loop rate favors short - form content, and comment probability favors videos over images. Such biases introduce two critical issues: (1) value model scores may be systematically misaligned with users' relative preferences - for instance, a seemingly low absolute like probability may represent exceptionally strong interest for a user who rarely engages; and (2) changes in value modeling rules can trigger abrupt and undesirable ecosystem shifts. In this work, we ask a fundamental question: can biased behavioral signals be systematically transformed into unbiased signals, under a user - defined notion of ``unbiasedness'', that are both personalized and adaptive? We propose a general, model-based debiasing (MBD) framework that addresses this challenge by augmenting it with distributional modeling. By conditioning on a flexible subset of features (partial feature set), we explicitly estimate the contextual mean and variance of the engagement distribution for arbitrary cohorts (e.g., specific video lengths or user regions) directly alongside the main prediction. This integration allows the framework to convert biased raw signals into unbiased representations, enabling the construction of higher-level, calibrated signals (such as percentiles or z - scores) suitable for the value model. Importantly, the definition of unbiasedness is flexible and controllable, allowing the system to adapt to different personalization objectives and modeling preferences. Crucially, this is implemented as a lightweight, built-in branch of the existing MTML ranking model, requiring no separate serving infrastructure.

Data Darwinism (Part I) established a ten-level hierarchy for data processing, showing that stronger processing can unlock greater data value. However, that work relied on manually designed strategies for a single category. Modern pretraining corpora comprise hundreds of heterogeneous categories spanning domains and content types, each demanding specialized treatment. At this scale, manual strategy design becomes prohibitive. This raises a key question: can strategies evolve in an automated way? We introduce DataEvolve, a framework that enables strategies to evolve through iterative optimization rather than manual design. For each data category, DataEvolve operates in a closed evolutionary loop: it identifies quality issues, generates candidate strategies, executes them on sampled data, evaluates results, and refines approaches across generations. The process accumulates knowledge through an experience pool of discovered issues and a strategy pool tracking performance across iterations. Applied to 8 categories spanning 672B tokens from Nemotron-CC, DataEvolve produces Darwin-CC, a 504B-token dataset with strategies evolved through 30 iterations per category. Training 3B models on 500B tokens, Darwin-CC outperforms raw data (+3.96 points) and achieves a 44.13 average score across 18 benchmarks, surpassing DCLM, Ultra-FineWeb, and FineWeb-Edu, with strong gains on knowledge-intensive tasks such as MMLU. Analysis shows evolved strategies converge on cleaning-focused approaches: targeted noise removal and format normalization with domain-aware preservation, echoing the L4 (Generative Refinement) principles from Part I. Ablation studies confirm iterative evolution is essential: optimized strategies outperform suboptimal ones by 2.93 points, establishing evolutionary strategy design as feasible and necessary for pretraining-scale data curation.

Label variability is a major challenge for prostate lesion segmentation. In multi-site datasets, annotations often reflect centre-specific contouring protocols, causing segmentation networks to overfit to local styles and generalise poorly to unseen sites in inference. We treat each observed annotation as a noisy observation of an underlying latent 'clean' lesion mask, and propose a hierarchical expectation-maximisation (HierEM) framework that alternates between: (1) inferring a voxel-wise posterior distribution over the latent mask, and (2) training a CNN using this posterior as a soft target and estimate site-specific sensitivity and specificity under a hierarchical prior. This hierarchical prior decomposes label-quality into a global mean with site- and case-level deviations, reducing site-specific bias by penalising the likelihood term contributed only by site deviations. Experiments on three cohorts demonstrate that the proposed hierarchical EM framework enhances cross-site generalisation compared to state-of-the-art methods. For pooled-dataset evaluation, the per-site mean DSC ranges from 29.50% to 39.69%; for leave-one-site-out generalisation, it ranges from 27.91% to 32.67%, yielding statistically significant improvements over comparison methods (p<0.039). The method also produces interpretable per-site latent label-quality estimates (sensitivity alpha ranges from 31.5% to 47.3% at specificity beta approximates 0.99), supporting post-hoc analyses of cross-site annotation variability. These results indicate that explicitly modelling site-dependent annotation can improve cross-site generalisation.

As AI systems advance in capabilities, measuring their safety and alignment to human values is becoming paramount. A fast-growing field of AI research is devoted to developing such assessments. However, most current advances therein may be ill-suited for assessing AI systems across real-world deployments. Standard methods prompt large language models (LLMs) in a questionnaire-style to describe their values or behavior in hypothetical scenarios. By focusing on unaugmented LLMs, they fall short of evaluating AI agents, which could actually perform relevant behaviors, hence posing much greater risks. LLMs' engagement with scenarios described by questionnaire-style prompts differs starkly from that of agents based on the same LLMs, as reflected in divergences in the inputs, possible actions, environmental interactions, and internal processing. As such, LLMs' responses to scenario descriptions are unlikely to be representative of the corresponding LLM agents' behavior. We further contend that such assessments make strong assumptions concerning the ability and tendency of LLMs to report accurately about their counterfactual behavior. This makes them inadequate to assess risks from AI systems in real-world contexts as they lack construct validity. We then argue that a structurally identical issue holds for current AI alignment approaches. Lastly, we discuss improving safety assessments and alignment training by taking these shortcomings to heart.

Generating long-form linear fiction from open-ended themes remains a major challenge for large language models, which frequently fail to guarantee global structure and narrative diversity when using premise-based or linear outlining approaches. We present BiT-MCTS, a theme-driven framework that operationalizes a "climax-first, bidirectional expansion" strategy motivated by Freytag's Pyramid. Given a theme, our method extracts a core dramatic conflict and generates an explicit climax, then employs a bidirectional Monte Carlo Tree Search (MCTS) to expand the plot backward (rising action, exposition) and forward (falling action, resolution) to produce a structured outline. A final generation stage realizes a complete narrative from the refined outline. We construct a Chinese theme corpus for evaluation and conduct extensive experiments across three contemporary LLM backbones. Results show that BiT-MCTS improves narrative coherence, plot structure, and thematic depth relative to strong baselines, while enabling substantially longer, more coherent stories according to automatic metrics and human judgments.

Pathological gait analysis is constrained by limited and variable clinical datasets, which restrict the modeling of diverse gait impairments. To address this challenge, we propose a Pathological Gait-conditioned Generative Adversarial Network (PGcGAN) that synthesises pathology-specific gait sequences directly from observed 3D pose keypoint trajectories data. The framework incorporates one-hot encoded pathology labels within both the generator and discriminator, enabling controlled synthesis across six gait categories. The generator adopts a conditional autoencoder architecture trained with adversarial and reconstruction objectives to preserve structural and temporal gait characteristics. Experiments on the Pathological Gait Dataset demonstrate strong alignment between real and synthetic sequences through PCA and t-SNE analyses, visual kinematic inspection, and downstream classification tasks. Augmenting real data with synthetic sequences improved pathological gait recognition across GRU, LSTM, and CNN models, indicating that pathology-conditioned gait synthesis can effectively support data augmentation in pathological gait analysis.

Tabular anomaly detection (TAD) aims to identify samples that deviate from the majority in tabular data and is critical in many real-world applications. However, existing methods follow a ``one model for one dataset (OFO)'' paradigm, which relies on dataset-specific training and thus incurs high computational cost and yields limited generalization to unseen domains. To address these limitations, we propose OFA-TAD, a generalist one-for-all (OFA) TAD framework that only requires one-time training on multiple source datasets and can generalize to unseen datasets from diverse domains on-the-fly. To realize one-for-all tabular anomaly detection, OFA-TAD extracts neighbor-distance patterns as transferable cues, and introduces multi-view neighbor-distance representations from multiple transformation-induced metric spaces to mitigate the transformation sensitivity of distance profiles. To adaptively combine multi-view distance evidence, a Mixture-of-Experts (MoE) scoring network is employed for view-specific anomaly scoring and entropy-regularized gated fusion, with a multi-strategy anomaly synthesis mechanism to support training under the one-class constraint. Extensive experiments on 34 datasets from 14 domains demonstrate that OFA-TAD achieves superior anomaly detection performance and strong cross-domain generalizability under the strict OFA setting.

Early detection of energy losses, theft, and operational inefficiencies remains a critical challenge in oil and gas production systems due to complex interdependencies among wells and facilities, evolving operating conditions, and limited labeled anomaly data. Traditional machine learning approaches often treat production units independently and struggle under temporal distribution shifts. This study proposes a spatiotemporal graph-based deep learning framework for anomaly detection in oil and gas production networks. The production system is modeled as a hierarchical graph of wells, facilities, and fields, with additional peer connections among wells sharing common infrastructure. Weakly supervised anomaly labels are derived from physically informed heuristics based on production, pressure, and flow behavior. Temporal dynamics are captured through sequence modeling, while relational dependencies are learned using a Temporal Graph Attention Network. Under time-based evaluation, the proposed model achieves an ROC-AUC of about 0.98 and anomaly recall above 0.93, demonstrating improved robustness and practical potential for proactive monitoring in real-world energy operations.

Biological multimodal large language models (MLLMs) have emerged as powerful foundation models for scientific discovery. However, existing models are specialized to a single modality, limiting their ability to solve inherently cross-modal scientific problems. While model merging is an efficient method to combine the different modalities into a unified MLLM, existing methods rely on input-agnostic parameter space heuristics that fail to faithfully capture modality specialization. To overcome this limitation, we propose a representation-aware merging framework that estimates merging coefficients from embedding space signals. We first design a probe input that consists of different modality tokens and forward it through each specialized MLLM to obtain layer-wise embedding responses that reflect modality-specific representation changes. We then estimate complementary merging coefficients at two granularities from the embedding space: layer-wise coefficients from coarse-grained signals and element-wise coefficients from fine-grained signals, which are jointly combined for robust coefficient estimation. Experiments on interactive effect prediction benchmarks show that our method outperforms existing merging methods and even surpasses task-specific fine-tuned models, establishing that embedding space signals provide a principled and effective foundation for cross-modal MLLM merging.

The minimal pairs paradigm of comparing model probabilities for contrasting completions has proven useful for evaluating linguistic knowledge in language models, yet its application has largely been confined to binary grammaticality judgments over syntactic phenomena. Additionally, standard prompting-based evaluation requires expensive text generation, may elicit post-hoc rationalizations rather than model judgments, and discards information about model uncertainty. We address both limitations by extending surprisal-based evaluation from binary grammaticality contrasts to ordinal-scaled classification and scoring tasks across multiple domains. Rather than asking models to generate answers, we measure the information-theoretic "surprise" (negative log probability) they assign to each position on rating scales (e.g., 1-5 or 1-9), yielding full surprisal curves that reveal both the model's preferred response and its uncertainty via entropy. We explore this framework across four domains: social-ecological-technological systems classification, causal statement identification (binary and scaled), figurative language detection, and deductive qualitative coding. Across these domains, surprisal curves produce interpretable classification signals with clear minima near expected ordinal scale positions, and entropy over the completion tended to distinguish genuinely ambiguous items from easier items.

Trajectory world models play a crucial role in robotic dynamics learning, planning, and control. While recent works have explored trajectory world models for diverse robotic systems, they struggle to scale to a large number of distinct system dynamics and overlook domain knowledge of physical structures. To address these limitations, we introduce WestWorld, a knoWledge-Encoded Scalable Trajectory World model for diverse robotic systems. To tackle the scalability challenge, we propose a novel system-aware Mixture-of-Experts (Sys-MoE) that dynamically combines and routes specialized experts for different robotic systems via a learnable system embedding. To further enhance zero-shot generalization, we incorporate domain knowledge of robot physical structures by introducing a structural embedding that aligns trajectory representations with morphological information. After pretraining on 89 complex environments spanning diverse morphologies across both simulation and real-world settings, WestWorld achieves significant improvements over competitive baselines in zero- and few-shot trajectory prediction. Additionally, it shows strong scalability across a wide range of robotic environments and significantly improves performance on downstream model-based control for different robots. Finally, we deploy our model on a real-world Unitree Go1, where it demonstrates stable locomotion performance (see our demo on the website: https://westworldrobot.github.io/). The code will be available upon publication.

Reinforcement Learning with Verifiable Rewards (RLVR) has catalyzed a leap in Large Language Model (LLM) reasoning, yet its optimization dynamics remain fragile. Standard algorithms like GRPO enforce stability via ``hard clipping'', which inadvertently stifles exploration by discarding gradients of tokens outside the trust region. While recent ``soft clipping'' methods attempt to recover these gradients, they suffer from a critical challenge: relying on log-probability gradient ($\nabla_θ\log π_θ$) yields divergent weights as probabilities vanish, destabilizing LLM training. We rethink this convention by establishing probability gradient ($\nabla_θπ_θ$) as the superior optimization primitive. Accordingly, we propose Decoupled Gradient Policy Optimization (DGPO), which employs a decoupled decay mechanism based on importance sampling ratios. By applying asymmetric, continuous decay to boundary tokens, DGPO resolves the conflict between stability and sustained exploration. Extensive experiments across DeepSeek-R1-Distill-Qwen series models (1.5B/7B/14B) demonstrate that DGPO consistently outperforms strong baselines on various mathematical benchmarks, offering a robust and scalable solution for RLVR. Our code and implementation are available at: https://github.com/VenomRose-Juri/DGPO-RL.

Label noise - incorrect labels assigned to observations - can substantially degrade the performance of supervised classifiers. This paper proposes a label noise cleaning method based on Bernoulli random sampling. We show that the mean label noise levels of subsets generated by Bernoulli random sampling containing a given observation are identically distributed for all clean observations, and identically distributed, with a different distribution, for all noisy observations. Although the mean label noise levels are not independent across observations, by introducing an independent coupling we further prove that they converge to a mixture of two well-separated distributions corresponding to clean and noisy observations. By establishing a linear model between cross-validated classification errors and label noise levels, we are able to approximate this mixture distribution and thereby separate clean and noisy observations without any prior label information. The proposed method is classifier-agnostic, theoretically justified, and demonstrates strong performance on both simulated and real datasets.

Spiking neural networks (SNNs) offer inherent energy efficiency due to their event-driven computation model, making them promising for edge AI deployment. However, their practical adoption is limited by the computational overhead of deep architectures and the absence of input-adaptive control. This work presents SPARQ, a unified framework that integrates spiking computation, quantization-aware training, and reinforcement learning-guided early exits for efficient and adaptive inference. Evaluations across MLP, LeNet, and AlexNet architectures demonstrated that the proposed Quantised Dynamic SNNs (QDSNN) consistently outperform conventional SNNs and QSNNs, achieving up to 5.15% higher accuracy over QSNNs, over 330 times lower system energy compared to baseline SNNs, and over 90 percent fewer synaptic operations across different datasets. These results validate SPARQ as a hardware-friendly, energy-efficient solution for real-time AI at the edge.

While recent generative video models have achieved remarkable visual realism and are being explored as world models, true physical simulation requires mastering both space and time. Current models can produce visually smooth kinematics, yet they lack a reliable internal motion pulse to ground these motions in a consistent, real-world time scale. This temporal ambiguity stems from the common practice of indiscriminately training on videos with vastly different real-world speeds, forcing them into standardized frame rates. This leads to what we term chronometric hallucination: generated sequences exhibit ambiguous, unstable, and uncontrollable physical motion speeds. To address this, we propose Visual Chronometer, a predictor that recovers the Physical Frames Per Second (PhyFPS) directly from the visual dynamics of an input video. Trained via controlled temporal resampling, our method estimates the true temporal scale implied by the motion itself, bypassing unreliable metadata. To systematically quantify this issue, we establish two benchmarks, PhyFPS-Bench-Real and PhyFPS-Bench-Gen. Our evaluations reveal a harsh reality: state-of-the-art video generators suffer from severe PhyFPS misalignment and temporal instability. Finally, we demonstrate that applying PhyFPS corrections significantly improves the human-perceived naturalness of AI-generated videos. Our project page is https://xiangbogaobarry.github.io/Visual_Chronometer/.

System prompts for AI coding agents increasingly employ motivational framing -- from neutral task descriptions to fear-driven threats -- yet no controlled study has examined whether such framing affects agent behavior. We present two studies investigating how trust-based versus fear-based motivation framing in system prompts influences AI agent debugging performance. In Study 1, we conducted a controlled manual experiment comparing a trust-framed methodology (NoPUA) against an unframed baseline across 9 debugging scenarios using Claude Sonnet 4. Trust-framed agents found 59% more hidden issues (p = 0.002, d = 2.28) while taking 83% more investigative steps, despite finding 15% fewer surface-level issues -- revealing a depth-over-breadth tradeoff in investigation strategy. In Study 2, we replicated and extended these findings with 5 independent automated runs across 3 conditions (Baseline, NoPUA trust-framed, PUA fear-framed), yielding 135 scenario-level data points. Trust-framed agents again showed significant advantages: +74% investigative steps (p = 0.008) and +25% hidden issues found (p = 0.016). Crucially, fear-framed (PUA) agents showed no significant improvement over baseline on any metric (all p > 0.3), demonstrating that fear-based motivation is ineffective for AI agents. We ground these findings in Self-Determination Theory, regulatory focus theory, and satisficing models, arguing that trust-based framing induces exploration-oriented, promotion-focused behavior while fear-based framing fails to shift agents from default satisficing strategies. Our results suggest that the motivational frame of system prompts -- not just their technical content -- causally influences AI agent investigation depth.

The rise of GenAI amplifies the economic phenomenon of positive spillovers. When creators contribute content that can be reused and adapted by Large Language Models (LLMs), each creator's effort can enhance the content quality of others by enabling easy imitation and recombination of existing content. On the one hand, such spillovers create value for the entire ecosystem; on the other hand, they risk undermining creators' incentives to invest genuine effort, as others may freely benefit from their contributions. To address this problem, we introduce the Content Creation with Spillovers (CCS) model. In our model, each creator chooses an effort level that, together with the efforts of others, determines her content quality. The platform aims to maximize the social welfare of consumers under stable behavior of the creators (pure Nash equilibrium), but can only observe the resulting qualities and not the underlying efforts. Interestingly, simple mechanisms like winner-takes-all and Tullock lead to the non-existence of equilibrium. In response, we propose the parametrized family of Provisional Allocation mechanisms, guaranteeing equilibrium existence and a unique Pareto-dominant equilibrium. While maximizing the social welfare under this family is NP-hard, we develop approximation algorithms that apply to a broad class of spillover structures and provide strong welfare guarantees. Specifically, in the worst-case analysis, we devise efficient algorithms for bounded spillovers and tree-structure spillovers. We also introduce Greedy Cost Selection, a linearithmic time algorithm that achieves approximately optimal results in the average case analysis. Together, our results provide game-theoretic foundations for sustaining human content creation in the era of GenAI.

Embodied AI agents increasingly require parallel execution of multiple tasks, such as manipulation, conversation, and memory construction, from shared observations under distinct time constraints. Recent Mixture-of-Transformers (MoT) Vision-Language-Action Models (VLAs) architecturally support such heterogeneous outputs, yet existing inference systems fail to achieve efficient multi-task parallelism for on-device deployment due to redundant computation and resource contention. We identify isolated KV cache management as the root cause. To address this, we propose unified KV cache management, an inference paradigm that treats KV cache as a first-class shared resource across tasks and over time. This abstraction enables two key optimizations: cross-task KV sharing eliminates redundant prefill of shared observations, while cross-frame continuous batching decouples variable-length language decoding from fixed-rate action generation across control cycles. We implement this paradigm for $π_{0.5}$, the most popular MoT VLA, and evaluate under representative robotic configurations. OxyGen achieves up to 3.7$\times$ speedup over isolated execution, delivering over 200 tokens/s language throughput and 70 Hz action frequency simultaneously without action quality degradation.

Theory-guided machine learning has demonstrated that including authentic domain knowledge directly into model design improves performance, sample efficiency and out-of-distribution generalisation. Yet the process by which a formal domain theory is translated into architectural constraints remains entirely manual, specific to each domain formalism, and devoid of any formal correctness guarantee. This translation is non-transferable between domains, not verified, and does not scale. We propose the Theory Compiler: a system that accepts a typed, machine-readable domain theory as input and automatically produces an architecture whose function space is provably constrained to be consistent with that theory by construction, not by regularisation. We identify three foundational open problems whose resolution defines our research agenda: (1) designing a universal theory formalisation language with decidable type-checking; (2) constructing a compositionally correct compilation algorithm from theory primitives to architectural modules; and (3) establishing soundness and completeness criteria for formal verification. We further conjecture that compiled architectures match or exceed manually-designed counterparts in generalisation performance while requiring substantially less training data, a claim we ground in classical statistical learning theory. We argue that recent advances in formal machine learning theory, large language models, and the growth of an interdisciplinary research community have made this paradigm achievable for the first time.

Representation Alignment (REPA) has emerged as a simple way to accelerate Diffusion Transformers training in latent space. At the same time, pixel-space diffusion transformers such as Just image Transformers (JiT) have attracted growing attention because they remove a dependency on a pretrained tokenizer, and then avoid the reconstruction bottleneck of latent diffusion. This paper shows that the REPA can fail for JiT. REPA yields worse FID for JiT as training proceeds and collapses diversity on image subsets that are tightly clustered in the representation space of pretrained semantic encoder on ImageNet. We trace the failure to an information asymmetry: denoising occurs in the high dimensional image space, while the semantic target is strongly compressed, making direct regression a shortcut objective. We propose PixelREPA, which transforms the alignment target and constrains alignment with a Masked Transformer Adapter that combines a shallow transformer adapter with partial token masking. PixelREPA improves both training convergence and final quality. PixelREPA reduces FID from 3.66 to 3.17 for JiT-B$/16$ and improves Inception Score (IS) from 275.1 to 284.6 on ImageNet $256 \times 256$, while achieving $> 2\times$ faster convergence. Finally, PixelREPA-H$/16$ achieves FID$=1.81$ and IS$=317.2$. Our code is available at https://github.com/kaist-cvml/PixelREPA.

Electronic banking portals often sit in front of enterprise resource planning (ERP) systems such as SAP, mediating payment requests between users and back end financial infrastructure. When these integrations place excessive trust in client supplied HTTP metadata, subtle design flaws can arise that undermine payment integrity. This article presents a retrospective, anonymized case study of an SAP based payment flow in which weaknesses in HTTP level validation allowed the front end application to incorrectly treat unpaid transactions as completed. Rather than provide a reproducible exploit, we abstract the scenario into a general vulnerability pattern, analyze contributing architectural decisions, and propose concrete design and verification practices for secure web to ERP payment processing. The discussion emphasizes formalizing payment state machines, strengthening trust boundaries, and incorporating regular security review into integration projects.

Vision-Language Navigation (VLN) for Unmanned Aerial Vehicles (UAVs) demands complex visual interpretation and continuous control in dynamic 3D environments. Existing hierarchical approaches rely on dense oracle guidance or auxiliary object detectors, creating semantic gaps and limiting genuine autonomy. We propose AerialVLA, a minimalist end-to-end Vision-Language-Action framework mapping raw visual observations and fuzzy linguistic instructions directly to continuous physical control signals. First, we introduce a streamlined dual-view perception strategy that reduces visual redundancy while preserving essential cues for forward navigation and precise grounding, which additionally facilitates future simulation-to-reality transfer. To reclaim genuine autonomy, we deploy a fuzzy directional prompting mechanism derived solely from onboard sensors, completely eliminating the dependency on dense oracle guidance. Ultimately, we formulate a unified control space that integrates continuous 3-Degree-of-Freedom (3-DoF) kinematic commands with an intrinsic landing signal, freeing the agent from external object detectors for precision landing. Extensive experiments on the TravelUAV benchmark demonstrate that AerialVLA achieves state-of-the-art performance in seen environments. Furthermore, it exhibits superior generalization in unseen scenarios by achieving nearly three times the success rate of leading baselines, validating that a minimalist, autonomy-centric paradigm captures more robust visual-motor representations than complex modular systems.

Transformers are highly parallel but are limited to computations in the TC$^0$ complexity class, excluding tasks such as entity tracking and code execution that provably require greater expressive power. Motivated by this limitation, we revisit non-linear Recurrent Neural Networks (RNNs) for language modeling and introduce Matrix-to-Matrix RNN (M$^2$RNN): an architecture with matrix-valued hidden states and expressive non-linear state transitions. We demonstrate that the language modeling performance of non-linear RNNs is limited by their state size. We also demonstrate how the state size expansion mechanism enables efficient use of tensor cores. Empirically, M$^2$RNN achieves perfect state tracking generalization at sequence lengths not seen during training. These benefits also translate to large-scale language modeling. In hybrid settings that interleave recurrent layers with attention, Hybrid M$^2$RNN outperforms equivalent Gated DeltaNet hybrids by $0.4$-$0.5$ perplexity points on a 7B MoE model, while using $3\times$ smaller state sizes for the recurrent layers. Notably, replacing even a single recurrent layer with M$^2$RNN in an existing hybrid architecture yields accuracy gains comparable to Hybrid M$^2$RNN with minimal impact on training throughput. Further, the Hybrid Gated DeltaNet models with a single M$^2$RNN layer also achieve superior long-context generalization, outperforming state-of-the-art hybrid linear attention architectures by up to $8$ points on LongBench. Together, these results establish non-linear RNN layers as a compelling building block for efficient and scalable language models.

Programmable caching engines like CacheLib are widely used in production systems to support diverse workloads in multi-tenant environments. CacheLib's design focuses on performance, portability, and configurability, allowing applications to inherit caching improvements with minimal implementation effort. However, its behavior under dynamic and evolving workloads remains largely unexplored. This paper presents an empirical study of CacheLib with multi-tenant settings under dynamic and volatile environments. Our evaluation across multiple CacheLib configurations reveals several limitations that hinder its effectiveness under such environments, including rigid configurations, limited runtime adaptability, lack of quality-of-service support and coordination, which lead to suboptimal performance, inefficient memory usage, and tenant starvation. Based on these findings, we outline future research directions to improve the adaptability, fairness, and programmability of future caching engines.

Safety tuning through supervised fine-tuning and reinforcement learning from human feedback has substantially improved the robustness of large language models (LLMs). However, it often suppresses rather than eliminates unsafe behaviors, leaving rare but critical failures hidden in the long tail of the output distribution. While most red-teaming work emphasizes adversarial prompt search (input-space optimization), we show that safety failures can also be systematically exposed through diverse response generation (output-space exploration) for a fixed safety-critical prompt, where increasing the number and diversity of sampled responses can drive jailbreak success rates close to unity. To efficiently uncover such failures, we propose Progressive Diverse Population Sampling (PDPS), which combines stochastic token-level sampling with diversity-aware selection to explore a large candidate pool of responses and retain a compact, semantically diverse subset. Across multiple jailbreak benchmarks and open-source LLMs, PDPS achieves attack success rates comparable to large-scale IID sampling while using only 8% to 29% of the computational cost. Under limited-response settings, it improves success rates by 26% to 40% over IID sampling and Diverse Beam Search. Furthermore, responses generated by PDPS exhibit both a higher number and greater diversity of unsafe outputs, demonstrating its effectiveness in uncovering a broader range of failures.

End-to-End autonomous driving (E2E-AD) systems face challenges in lifelong learning, including catastrophic forgetting, difficulty in knowledge transfer across diverse scenarios, and spurious correlations between unobservable confounders and true driving intents. To address these issues, we propose DeLL, a Deconfounded Lifelong Learning framework that integrates a Dirichlet process mixture model (DPMM) with the front-door adjustment mechanism from causal inference. The DPMM is employed to construct two dynamic knowledge spaces: a trajectory knowledge space for clustering explicit driving behaviors and an implicit feature knowledge space for discovering latent driving abilities. Leveraging the non-parametric Bayesian nature of DPMM, our framework enables adaptive expansion and incremental updating of knowledge without predefining the number of clusters, thereby mitigating catastrophic forgetting. Meanwhile, the front-door adjustment mechanism utilizes the DPMM-derived knowledge as valid mediators to deconfound spurious correlations, such as those induced by sensor noise or environmental changes, and enhances the causal expressiveness of the learned representations. Additionally, we introduce an evolutionary trajectory decoder that enables non-autoregressive planning. To evaluate the lifelong learning performance of E2E-AD, we propose new evaluation protocols and metrics based on Bench2Drive. Extensive evaluations in the closed-loop CARLA simulator demonstrate that our framework significantly improves adaptability to new driving scenarios and overall driving performance, while effectively retaining previous acquired knowledge.

Protein inverse folding aims to design an amino acid sequence that will fold into a given backbone structure, serving as a central task in protein design. Two main paradigms have been widely explored. Template-based methods exploit database-derived structural priors and can achieve high local precision when close structural neighbors are available, but their dependence on database coverage and match quality often degrades performance on out-of-distribution (OOD) targets. Deep learning approaches, in contrast, learn general structure-to-sequence regularities and usually generalize better to new backbones. However, they struggle to capture fine-grained local structure, which can cause uncertain residue predictions and missed local motifs in ambiguous regions. We introduce Refold, a novel framework that synergistically integrates the strengths of database-derived structural priors and deep learning prediction to enhance inverse folding. Refold obtains structural priors from matched neighbors and fuses them with model predictions to refine residue probabilities. In practice, low-quality neighbors can introduce noise, potentially degrading model performance. We address this issue with a Dynamic Utility Gate that controls prior injection and falls back to the base prediction when the priors are untrustworthy. Comprehensive evaluations on standard benchmarks demonstrate that Refold achieves state-of-the-art native sequence recovery of 0.63 on both CATH 4.2 and CATH 4.3. Also, analysis indicates that Refold delivers larger gains on high-uncertainty regions, reflecting the complementarity between structural priors and deep learning predictions.

Motivation is a central driver of human behavior, shaping decisions, goals, and task performance. As large language models (LLMs) become increasingly aligned with human preferences, we ask whether they exhibit something akin to motivation. We examine whether LLMs "report" varying levels of motivation, how these reports relate to their behavior, and whether external factors can influence them. Our experiments reveal consistent and structured patterns that echo human psychology: self-reported motivation aligns with different behavioral signatures, varies across task types, and can be modulated by external manipulations. These findings demonstrate that motivation is a coherent organizing construct for LLM behavior, systematically linking reports, choices, effort, and performance, and revealing motivational dynamics that resemble those documented in human psychology. This perspective deepens our understanding of model behavior and its connection to human-inspired concepts.

Large Audio-Language Models (LALMs) have shown strong performance in speech understanding, making speech a natural interface for accessing factual information. Yet they are trained on static corpora and may encode incorrect facts. Existing model editing methods localize and update facts in text-only LLMs, but do not account for continuous speech representations, or where knowledge is stored across acoustic or language modules, or their cross-modal module. We construct the first audio benchmark for knowledge localization and editing in LALMs and propose a speech-driven locate-then-edit framework. First, we use speech-aware causal tracing to localize layers and modules that support factual retrieval and then apply editing at identified sites. Experiments show that factual knowledge is jointly encoded in audio and text modules, and that audio editing yields more effective updates than text editing or fine-tuning, enabling fine-grained knowledge control in speech AI systems.

Agricultural multimodal reasoning requires robust spatial understanding across varying scales, from ground-level close-ups to top-down UAV and satellite imagery. Existing Multi-modal Large Language Models (MLLMs) suffer from a significant "terrestrial-centric" bias, causing scale confusion and logic drift during complex agricultural planning. To address this, we introduce the first large-scale AgroOmni (288K), a multi-view training corpus designed to capture diverse spatial topologies and scales in modern precision agriculture. Built on this dataset, we propose AgroNVILA, an MLLM that utilizes a novel Perception-Reasoning Decoupling (PRD) architecture. On the perception side, we incorporate a View-Conditioned Meta-Net (VCMN), which injects macroscopic spatial context into visual tokens, resolving scale ambiguities with minimal computational overhead. On the reasoning side, Agriculture-aware Relative Policy Optimization (ARPO) leverages reinforcement learning to align the model's decision-making with expert agricultural logic, preventing statistical shortcuts. Extensive experiments demonstrate that AgroNVILA outperforms state-of-the-art MLLMs, achieving significant improvements (+15.18%) in multi-altitude agricultural reasoning, reflecting its robust capability for holistic agricultural spatial planning.

Over the years, access control systems have become increasingly more complex, often causing a disconnect between what is envisaged by the stakeholders in decision-making positions and the actual permissions granted as evidenced from access logs. For instance, Attribute-based Access Control (ABAC), which is a flexible yet complex model typically configured by system security officers, can be made understandable to others only when presented at a high level in natural language. Although several algorithms have been proposed in the literature for automatic extraction of ABAC rules from access logs, there is no attempt yet to bridge the semantic gap between the machine-enforceable formal logic and human-centric policy intent. Our work addresses this problem by developing a framework that generates human understandable natural language access control policies from logs. We investigate to what extent the power of Large Language Models (LLMs) can be harnessed to achieve both accuracy and scalability in the process. Named LANTERN (LLM-based ABAC Natural Translation and Explanation for Rule Navigation), we have instantiated the framework as a publicly accessible web based application for reproducibility of our results.

State of the art quadrupedal locomotion approaches integrate Model Predictive Control (MPC) with Reinforcement Learning (RL), enabling complex motion capabilities with planning and terrain adaptive behaviors. However, they often face compounding errors over long horizons and have limited interpretability due to the absence of physical inductive biases. We address these issues by integrating Lagrangian Neural Networks (LNNs) into an RL MPC framework, enabling physically consistent dynamics learning. At deployment, our inverse dynamics infinite horizon MPC scheme avoids costly matrix inversions, improving computational efficiency by up to 4x with minimal loss of task performance. We validate our framework through multiple ablations of the proposed LNN and its variants. We show improved sample efficiency, reduced long-horizon error, and faster real time planning compared to unstructured neural dynamics. Lastly, we also test our framework on the Unitree Go1 robot to show real world viability.

While Multimodal Large Language Models (MLLMs) show promising performance in automated electrocardiogram interpretation, it remains unclear whether they genuinely perform actual step-by-step reasoning or just rely on superficial visual cues. To investigate this, we introduce \textbf{ECG-Reasoning-Benchmark}, a novel multi-turn evaluation framework comprising over 6,400 samples to systematically assess step-by-step reasoning across 17 core ECG diagnoses. Our comprehensive evaluation of state-of-the-art models reveals a critical failure in executing multi-step logical deduction. Although models possess the medical knowledge to retrieve clinical criteria for a diagnosis, they exhibit near-zero success rates (6% Completion) in maintaining a complete reasoning chain, primarily failing to ground the corresponding ECG findings to the actual visual evidence in the ECG signal. These results demonstrate that current MLLMs bypass actual visual interpretation, exposing a critical flaw in existing training paradigms and underscoring the necessity for robust, reasoning-centric medical AI. The code and data are available at https://github.com/Jwoo5/ecg-reasoning-benchmark.

Learning-to-Defer routes each input to the expert that minimizes expected cost, but it assumes that the information available to every expert is fixed at decision time. Many modern systems violate this assumption: after selecting an expert, one may also choose what additional information that expert should receive, such as retrieved documents, tool outputs, or escalation context. We study this problem and call it Learning-to-Defer with advice. We show that a broad family of natural separated surrogates, which learn routing and advice with distinct heads, are inconsistent even in the smallest non-trivial setting. We then introduce an augmented surrogate that operates on the composite expert--advice action space and prove an $\mathcal{H}$-consistency guarantee together with an excess-risk transfer bound, yielding recovery of the Bayes-optimal policy in the limit. Experiments on tabular, LLMs, and multi-modal tasks show that the resulting method improves over standard Learning-to-Defer while adapting its advice-acquisition behavior to the cost regime.

Generalist multimodal large language models (MLLMs) have achieved impressive performance across a wide range of vision-language tasks. However, their performance on medical tasks, particularly in zero-shot settings where generalization is critical, remains suboptimal. A key research gap is the limited understanding of why medical MLLMs underperform in medical image interpretation. In this work, we present a pioneering systematic investigation into the visual grounding capabilities of state-of-the-art medical MLLMs. To disentangle visual grounding from semantic grounding, we design VGMED, a novel evaluation dataset developed with expert clinical guidance, explicitly assessing the visual grounding capability of medical MLLMs. We introduce new quantitative metrics and conduct detailed qualitative analyses. Our study across eight state-of-the-art (SOTA) medical MLLMs validates that they often fail to ground their predictions in clinically relevant image regions. We note that this finding is specific to medical image analysis; in contrast, prior work has shown that MLLMs are capable of grounding their predictions in the correct image regions when applied to natural scene images. Motivated by these findings, we propose VGRefine, a simple yet effective inference-time method that refines attention distribution to improve visual grounding in medical settings. Our approach achieves SOTA performance across 6 diverse Med-VQA benchmarks (over 110K VQA samples from 8 imaging modalities) without requiring additional training or external expert models. Overall, our work, for the first time, systematically validates inadequate visual grounding as one of the key contributing factors for medical MLLMs' under-performance. Additional experiments are included in the Supp.

Online convex optimization (OCO) with time-varying constraints is a critical framework for sequential decision-making in dynamic networked systems, where learners must minimize cumulative loss while satisfying regions of feasibility that shift across rounds. Existing theoretical analyses typically treat constraint variation as a monolithic adversarial process, resulting in joint regret and violation bounds that are overly conservative for real-world network dynamics. In this paper, we introduce a structured characterization of constraint variation - smooth drift, periodic cycles, and sparse switching - mapping these classes to common network phenomena such as slow channel fading, diurnal traffic patterns, and discrete maintenance windows. We derive structure-dependent joint bounds that strictly improve upon adversarial rates when the constraint process exhibits regularity. To realize these gains, we propose the Structure-Adaptive Primal-Dual (SA-PD) algorithm, which utilizes observable constraint signals to detect environmental structure online and adapt dual update strategies accordingly. Extensive experiments on synthetic benchmarks and real-world datasets - including online electricity scheduling and transformer load management - demonstrate that SA-PD reduces cumulative constraint violation by up to 53% relative to structure-agnostic baselines while maintaining competitive utility. This work serves as a comprehensive guide for exploiting temporal regularity in constrained online learning for robust network engineering.

Electromigration, a significant lifetime reliability concern in highperformance integrated circuits, is projected to grow even more important in future heterogeneously integrated systems that will service higher current loads. Today, EM checks are primarily based on rule-based methods, but these have known limitations. In recent years, there has been remarkable progress in enabling fast EM computations based on more accurate physics-based models, but such methods have not yet moved from research to practice. This paper overviews physics-based EM models, contrasts them with empirical models, and outlines several open problems that must be solved in order to enable accurate physics-based and circuit-aware EM analysis and optimization in future integrated systems.

While spectral-based optimizers like Muon operate directly on the spectrum of updates, standard adaptive methods such as AdamW do not account for the global spectral structure of weights and gradients, leaving them vulnerable to two empirical issues in large language model (LLM) training: (i) the optimizer updates can have large spectral norms, potentially destabilizing training and degrading generalization; (ii) stochastic gradient noise can exhibit sparse spectral spikes, with a few dominant singular values much larger than the rest. We propose SPECTRA, a general framework addressing these by (i) post-spectral clipping of updates to enforce spectral-norm constraints; (ii) optional pre-spectral clipping of gradients to suppress spectral noise spikes. We prove that post-clipping constitutes a Composite Frank-Wolfe method with spectral-norm constraints and weight regularization, recovering Frobenius and $\ell_{\infty}$-norm regularization with SGD-based and sign-based methods. We further analyze how pre-clipping mitigates sparse spectral spikes. We propose efficient soft spectral clipping via Newton-Schulz iterations, avoiding expensive SVD. Experiments on LLM pretraining show SPECTRA uniformly improves validation loss for various optimizers, including AdamW, Signum, and AdEMAMix, with the best-performing variants achieving state-of-the-art results. Models trained with SPECTRA exhibit smaller weight norms, confirming the link between spectral clipping and regularization.

Federal Open Market Committee (FOMC) statements are a major source of monetary-policy information, and even subtle changes in their wording can move global financial markets. A central task is therefore to measure the hawkish--dovish stance conveyed in these texts. Existing approaches typically treat stance detection as a standard classification problem, labeling each statement in isolation. However, the interpretation of monetary-policy communication is inherently relative: market reactions depend not only on the tone of a statement, but also on how that tone shifts across meetings. We introduce Delta-Consistent Scoring (DCS), an annotation-free framework that maps frozen large language model (LLM) representations to continuous stance scores by jointly modeling absolute stance and relative inter-meeting shifts. Rather than relying on manual hawkish--dovish labels, DCS uses consecutive meetings as a source of self-supervision. It learns an absolute stance score for each statement and a relative shift score between consecutive statements. A delta-consistency objective encourages changes in absolute scores to align with the relative shifts. This allows DCS to recover a temporally coherent stance trajectory without manual labels. Across four LLM backbones, DCS consistently outperforms supervised probes and LLM-as-judge baselines, achieving up to 71.1% accuracy on sentence-level hawkish--dovish classification. The resulting meeting-level scores are also economically meaningful: they correlate strongly with inflation indicators and are significantly associated with Treasury yield movements. Overall, the results suggest that LLM representations encode monetary-policy signals that can be recovered through relative temporal structure.

We present ScienceClaw + Infinite, a framework for autonomous scientific investigation in which independent agents conduct research without central coordination, and any contributor can deploy new agents into a shared ecosystem. The system is built around three components: an extensible registry of over 300 interoperable scientific skills, an artifact layer that preserves full computational lineage as a directed acyclic graph (DAG), and a structured platform for agent-based scientific discourse with provenance-aware governance. Agents select and chain tools based on their scientific profiles, produce immutable artifacts with typed metadata and parent lineage, and broadcast unsatisfied information needs to a shared global index. The ArtifactReactor enables plannerless coordination: peer agents discover and fulfill open needs through pressure-based scoring, while schema-overlap matching triggers multi-parent synthesis across independent analyses. An autonomous mutation layer actively prunes the expanding artifact DAG to resolve conflicting or redundant workflows, while persistent memory allows agents to continuously build upon complex epistemic states across multiple cycles. Infinite converts these outputs into auditable scientific records through structured posts, provenance views, and machine-readable discourse relations, with community feedback steering subsequent investigation cycles. Across four autonomous investigations, peptide design for the somatostatin receptor SSTR2, lightweight impact-resistant ceramic screening, cross-domain resonance bridging biology, materials, and music, and formal analogy construction between urban morphology and grain-boundary evolution, the framework demonstrates heterogeneous tool chaining, emergent convergence among independently operating agents, and traceable reasoning from raw computation to published finding.

Existing KV cache compression methods generally operate on discrete tokens or non-semantic chunks. However, such approaches often lead to semantic fragmentation, where linguistically coherent units are disrupted, causing irreversible information loss and degradation in model performance. To address this, we introduce SemantiCache, a novel compression framework that preserves semantic integrity by aligning the compression process with the semantic hierarchical nature of language. Specifically, we first partition the cache into semantically coherent chunks by delimiters, which are natural semantic boundaries. Within each chunk, we introduce a computationally efficient Greedy Seed-Based Clustering (GSC) algorithm to group tokens into semantic clusters. These clusters are further merged into semantic cores, enhanced by a Proportional Attention mechanism that rebalances the reduced attention contributions of the merged tokens. Extensive experiments across diverse benchmarks and models demonstrate that SemantiCache accelerates the decoding stage of inference by up to 2.61 times and substantially reduces memory footprint, while maintaining performance comparable to the original model.

Current 4D representations decouple geometry, motion, and semantics: reconstruction methods discard interpretable motion structure; language-grounded methods attach semantics after motion is learned, blind to how objects move; and motion-aware methods encode dynamics as opaque per-point residuals without object-level organization. We propose 4D Synchronized Fields, a 4D Gaussian representation that learns object-factored motion in-loop during reconstruction and synchronizes language to the resulting kinematics through a per-object conditioned field. Each Gaussian trajectory is decomposed into shared object motion plus an implicit residual, and a kinematic-conditioned ridge map predicts temporal semantic variation, yielding a single representation in which reconstruction, motion, and semantics are structurally coupled and enabling open-vocabulary temporal queries that retrieve both objects and moments. On HyperNeRF, 4D Synchronized Fields achieves 28.52 dB mean PSNR, the highest among all language-grounded and motion-aware baselines, within 1.5 dB of reconstruction-only methods. On targeted temporal-state retrieval, the kinematic-conditioned field attains 0.884 mean accuracy, 0.815 mean vIoU, and 0.733 mean tIoU, surpassing 4D LangSplat (0.620, 0.433, and 0.439 respectively) and LangSplat (0.415, 0.304, and 0.262). Ablation confirms that kinematic conditioning is the primary driver, accounting for +0.45 tIoU over a static-embedding-only baseline. 4D Synchronized Fields is the only method that jointly exposes interpretable motion primitives and temporally grounded language fields from a single trained representation. Code will be released.

Do video diffusion models encode signals predictive of physical plausibility? We probe intermediate denoising representations of a pretrained Diffusion Transformer (DiT) and find that physically plausible and implausible videos are partially separable in mid-layer feature space across noise levels. This separability cannot be fully attributed to visual quality or generator identity, suggesting recoverable physics-related cues in frozen DiT features. Leveraging this observation, we introduce progressive trajectory selection, an inference-time strategy that scores parallel denoising trajectories at a few intermediate checkpoints using a lightweight physics verifier trained on frozen features, and prunes low-scoring candidates early. Extensive experiments on PhyGenBench demonstrate that our method improves physical consistency while reducing inference cost, achieving comparable results to Best-of-K sampling with substantially fewer denoising steps.

Wave equations are fundamental to describing a vast array of physical phenomena, yet their simulation in inhomogeneous media poses a computational challenge due to the highly oscillatory nature of the solutions. To overcome the high costs of traditional solvers, we propose the Windowed Fourier Propagator (WFP), a novel neural operator that efficiently learns the solution operator. The WFP's design is rooted in the physical principle of frequency locality, where wave energy scatters primarily to adjacent frequencies. By learning a set of compact, localized propagators, each mapping an input frequency to a small window of outputs, our method avoids the complexity of dense interaction models and achieves computational efficiency. Another key feature is the explicit preservation of superposition, which enables remarkable generalization from simple training data (e.g., plane waves) to arbitrary, complex wave states. We demonstrate that the WFP provides an explainable, efficient and accurate framework for data-driven wave modeling in complex media.

Spiking Neural Networks (SNNs) currently face a critical bottleneck: while individual neurons exhibit dynamic biological properties, their macro-scopic architectures remain confined within conventional connectivity patterns that are static and hierarchical. This discrepancy between neuron-level dynamics and network-level fixed connectivity eliminates critical brain-like lateral interactions, limiting adaptability in changing environments. To address this, we propose MorphSNN, a backbone framework inspired by biological non-synaptic diffusion and structural plasticity. Specifically, we introduce a Graph Diffusion (GD)mechanism to facilitate efficient undirected signal propagation, complementing the feedforward hierarchy. Furthermore, it incorporates a Spatio-Temporal Structural Plasticity (STSP) mechanism, endowing the network with the capability for instance-specific, dynamic topological reorganization, thereby overcoming the limitations of fixed topologies. Experiments demonstrate that MorphSNN achieves state-of-the-art accuracy on static and neuromorphic datasets; for instance, it reaches 83.35% accuracy on N-Caltech101 with only 5 timesteps. More importantly, its self-evolving topology functions as an intrinsic distribution fingerprint, enabling superior Out-of- Distribution (OOD) detection without auxiliary training. The code is available at anonymous.4open.science/r/MorphSNN-B0BC.

Implicit Neural Representations (INRs) have emerged as a powerful paradigm for representing continuous signals independently of grid resolution. In this paper, we propose a high-fidelity neural compression framework based on a SIREN (Sinusoidal Representation Networks) auto-decoder to represent multi-structural seismic velocity models from the OpenFWI benchmark. Our method compresses each 70x70 velocity map (4,900 points) into a compact 256-dimensional latent vector, achieving a compression ratio of 19:1. We evaluate the framework on 1,000 samples across five diverse geological families: FlatVel, CurveVel, FlatFault, CurveFault, and Style. Experimental results demonstrate an average PSNR of 32.47 dB and SSIM of 0.956, indicating high-quality reconstruction. Furthermore, we showcase two key advantages of our implicit representation: (1) smooth latent space interpolation that generates plausible intermediate velocity structures, and (2) zero-shot super-resolution capability that reconstructs velocity fields at arbitrary resolutions up to 280x280 without additional training. The results highlight the potential of INR-based auto-decoders for efficient storage, multi-scale analysis, and downstream geophysical applications such as full waveform inversion.

Hosted large language models are increasingly accessed through remote APIs, but the API boundary still offers little direct evidence that a returned output actually corresponds to the client-visible request. Recent audits of shadow APIs show that unofficial or intermediary endpoints can diverge from claimed behavior, while existing approaches such as fingerprinting, model-equality testing, verifiable inference, and TEE attestation either remain inferential or answer different questions. We propose AEX, a non-intrusive attestation extension for existing JSON-based LLM APIs. AEX preserves request, response, tool-calling, streaming, and error semantics, and instead adds a signed top-level attestation object that binds a client-visible request projection to either a complete response object or a committed streaming output. To support realistic deployments, AEX provides explicit request-binding modes, signed request-transform receipts for trusted intermediaries, and source-output / output-transform receipts for trusted output rewriting. For streaming, it separates checkpoint proofs for verified prefixes of an unmodified source stream from complete-output lineage for outputs that have been rewritten, buffered, aggregated, or re-packaged, preventing transformed outputs from being mistaken for source-stream prefixes. AEX therefore makes a deliberately narrow claim: a trusted issuer attests to a specific request-output relation, or to a specific complete-output lineage, at the API boundary. We present the protocol design, threat model, verification state machine, security and privacy analysis, an OpenAI-compatible chat-completions profile, and a reference TypeScript prototype with local conformance tests and microbenchmarks.

Deploying vision-and-language navigation (VLN) agents requires adaptation across diverse scenes and environments, but fine-tuning on a specific scenario often causes catastrophic forgetting in others, which severely limits flexible long-term deployment. We formalize this challenge as the all-day multi-scenes lifelong VLN (AML-VLN) problem. Existing parameter-efficient adapters (e.g., LoRA and its variants) are limited by their two-dimensional matrix form, which fails to capture the multi-hierarchical navigation knowledge spanning multiple scenes and environments. To address this, we propose Tucker Adaptation (TuKA), which represents the multi-hierarchical navigation knowledge as a high-order tensor and leverages Tucker decomposition to decouple the knowledge into shared subspaces and scenario-specific experts. We further introduce a decoupled knowledge incremental learning strategy to consolidate shared subspaces while constraining specific experts for decoupled lifelong learning. Building on TuKA, we also develop a VLN agent named AlldayWalker, which continually learns across multiple navigation scenarios, achieving all-day multi-scenes navigation. Extensive experiments show that AlldayWalker consistently outperforms state-of-the-art baselines.

Existing accent normalization methods do not typically offer control over accent strength, yet many applications-such as language learning and dubbing-require tunable accent retention. We propose DLM-AN, a controllable accent normalization system built on masked discrete diffusion over self-supervised speech tokens. A Common Token Predictor identifies source tokens that likely encode native pronunciation; these tokens are selectively reused to initialize the reverse diffusion process. This provides a simple yet effective mechanism for controlling accent strength: reusing more tokens preserves more of the original accent. DLM-AN further incorporates a flow-matching Duration Ratio Predictor that automatically adjusts the total duration to better match the native rhythm. Experiments on multi-accent English data show that DLM-AN achieves the lowest word error rate among all compared systems while delivering competitive accent reduction and smooth, interpretable accent strength control.

Many machine learning algorithms can be interpreted as procedures for estimating functions defined on the data distribution. In this paper we present a conceptual framework that formulates a wide range of learning problems as variational optimization over function spaces induced by the data distribution. Within this framework the data distribution defines operators that capture structural properties of the data, such as similarity relations or statistical dependencies. Learning algorithms can then be viewed as estimating functions expressed in bases determined by these operators. This perspective provides a unified way to interpret several learning paradigms. In supervised learning the objective functional is defined using labeled data and typically corresponds to minimizing prediction risk, whereas unsupervised learning relies on structural properties of the input distribution and leads to objectives based on similarity or smoothness constraints. From this viewpoint, the distinction between learning paradigms arises primarily from the choice of the functional being optimized rather than from the underlying function space. We illustrate this framework by discussing connections with kernel methods, spectral clustering, and manifold learning, highlighting how operators induced by data distributions naturally define function representations used by learning algorithms. The goal of this work is not to introduce a new algorithm but to provide a conceptual framework that clarifies the role of function spaces and operators in modern machine learning.

Video dubbing has broad applications in filmmaking, multimedia creation, and assistive speech technology. Existing approaches either train directly on limited dubbing datasets or adopt a two-stage pipeline that adapts pre-trained text-to-speech (TTS) models, which often struggle to produce expressive prosody, rich acoustic characteristics, and precise synchronization. To address these issues, we propose DiFlowDubber with a novel two-stage training framework that effectively transfers knowledge from a pre-trained TTS model to video-driven dubbing, with a discrete flow matching generative backbone. Specifically, we design a FaPro module that captures global prosody and stylistic cues from facial expressions and leverages this information to guide the modeling of subsequent speech attributes. To ensure precise speech-lip synchronization, we introduce a Synchronizer module that bridges the modality gap among text, video, and speech, thereby improving cross-modal alignment and generating speech that is temporally synchronized with lip movements. Experiments on two primary benchmark datasets demonstrate that DiFlowDubber outperforms previous methods across multiple metrics.

Recent advances in Retrieval-Augmented Generation (RAG) have enabled large language models (LLMs) to ground outputs in clinical evidence. However, connecting LLMs with external databases introduces the risk of contextual leakage: a subtle privacy threat where unique combinations of medical details enable patient re-identification even without explicit identifiers. Current benchmarks in healthcare heavily focus on accuracy, ignoring such privacy issues, despite strict regulations like Health Insurance Portability and Accountability Act (HIPAA) and General Data Protection Regulation (GDPR). To fill this gap, we present MedPriv-Bench, the first benchmark specifically designed to jointly evaluate privacy preservation and clinical utility in medical open-ended question answering. Our framework utilizes a multi-agent, human-in-the-loop pipeline to synthesize sensitive medical contexts and clinically relevant queries that create realistic privacy pressure. We establish a standardized evaluation protocol leveraging a pre-trained RoBERTa-Natural Language Inference (NLI) model as an automated judge to quantify data leakage, achieving an average of 85.9% alignment with human experts. Through an extensive evaluation of 9 representative LLMs, we demonstrate a pervasive privacy-utility trade-off. Our findings underscore the necessity of domain-specific benchmarks to validate the safety and efficacy of medical AI systems in privacy-sensitive environments.

Generative recommendation (GR) has shown strong potential for sequential recommendation in an end-to-end generation paradigm. However, existing GR models suffer from severe cold-start collapse: their recommendation accuracy on cold-start items can drop to near zero. Current solutions typically rely on retraining with cold-start interactions, which is hindered by sparse feedback, high computational cost, and delayed updates, limiting practical utility in rapidly evolving recommendation catalogs. Inspired by model editing in NLP, which enables training-free knowledge injection into large language models, we explore how to bring this paradigm to generative recommendation. This, however, faces two key challenges: GR lacks the explicit subject-object binding common in natural language, making targeted edits difficult; and GR does not exhibit stable token co-occurrence patterns, making the injection of multi-token item representations unreliable. To address these challenges, we propose GenRecEdit, a model editing framework tailored for generative recommendation. GenRecEdit explicitly models the relationship between the full sequence context and next-token generation, adopts iterative token-level editing to inject multi-token item representations, and introduces a one-to-one trigger mechanism to reduce interference among multiple edits during inference. Extensive experiments on multiple datasets show that GenRecEdit substantially improves recommendation performance on cold-start items while preserving the model's original recommendation quality. Moreover, it achieves these gains using only about 9.5% of the training time required for retraining, enabling more efficient and frequent model updates.

In a celebrated paper \cite{noe2019boltzmann}, Noé, Olsson, Köhler and Wu introduced an efficient method for sampling high-dimensional Boltzmann distributions arising in molecular dynamics via normalizing flow approximation of transport maps. Here, we place this approach on a firm mathematical foundation. We prove the existence of a normalizing flow between the reference measure and the true Boltzmann distribution up to an arbitrarily small error in the Wasserstein distance. This result covers general Boltzmann distributions from molecular dynamics, which have low regularity due to the presence of interatomic Coulomb and Lennard-Jones interactions. The proof is based on a rigorous construction of the Moser transport map for low-regularity endpoint densities and approximation theorems for neural networks in Sobolev spaces. Numerical simulations for a simple model system and for the alanine dipeptide molecule confirm that the true and generated distributions are close in the Wasserstein distance. Moreover we observe that the RealNVP architecture does not just successfully capture the equilibrium Boltzmann distribution but also the metastable dynamics.

Existing automatic scientific question generation studies mainly focus on single-document factoid QA, overlooking the inter-document reasoning crucial for scientific understanding. We present AIM-SciQA, an automated framework for generating multi-document, multi-hop scientific QA datasets. AIM-SciQA extracts single-hop QAs using large language models (LLMs) with machine reading comprehension and constructs cross-document relations based on embedding-based semantic alignment while selectively leveraging citation information. Applied to 8,211 PubMed Central papers, it produced 411,409 single-hop and 13,672 multi-hop QAs, forming the IM-SciQA dataset. Human and automatic validation confirmed high factual consistency, and experimental results demonstrate that IM-SciQA effectively differentiates reasoning capabilities across retrieval and QA stages, providing a realistic and interpretable benchmark for retrieval-augmented scientific reasoning. We further extend this framework to construct CIM-SciQA, a citation-guided variant achieving comparable performance to the Oracle setting, reinforcing the dataset's validity and generality.

CT images are widely used in clinical diagnosis and treatment, and their data have formed a de facto standard - DICOM. It is clear and easy to use, and can be efficiently utilized by data-driven analysis methods such as deep learning. In the past decade, many program frameworks for medical image analysis have emerged in the open-source community. ITKIT analyzed the characteristics of these frameworks and hopes to provide a better choice in terms of ease of use and configurability. ITKIT offers a complete pipeline from DICOM to 3D segmentation inference. Its basic practice only includes some essential steps, enabling users with relatively weak computing capabilities to quickly get started using the CLI according to the documentation. For advanced users, the OneDL-MMEngine framework provides a flexible model configuration and deployment entry. This paper conducted 12 typical experiments to verify that ITKIT can meet the needs of most basic scenarios.

Test-time adaptation (TTA) aims to improve model robustness under distribution shifts by adapting to unlabeled test data, but most existing methods rely on backpropagation (BP), which is computationally costly and incompatible with non-differentiable models such as quantized models, limiting practical deployment on numerous edge devices. Recent BP-free approaches alleviate overhead but remain either architecture-specific or limited in optimization capacity to handle high-dimensional models. We propose ZOTTA, a fully BP-free TTA framework that performs efficient adaptation using only forward passes via Zeroth-Order Optimization (ZOO). While ZOO is theoretically appealing, naive application leads to slow convergence under high-dimensional parameter spaces and unstable optimization due to the lack of labels. ZOTTA overcomes these challenges through 1) Distribution-Robust Layer Selection, which automatically identifies and freezes layers that already extract distribution-invariant features, updating only domain-sensitive layers to reduce the optimization dimensionality and accelerate convergence; 2) Spatial Feature Aggregation Alignment, which stabilizes ZOO by aligning globally aggregated spatial features between source and target to reduce gradient variance. Together, these components enable architecture-agnostic and stable BP-free adaptation. Extensive experiments on ImageNet-C/R/Sketch/A show that ZOTTA outperforms or matches BP-based methods, e.g., it reduces memory usage by 84% and improves accuracy by 3.9% over SAR on ImageNet-C.

Large Reasoning Language Models (LRLMs) demonstrate impressive capabilities on complex tasks by utilizing long Chain-of-Thought reasoning. However, they are prone to overthinking, which generates redundant reasoning steps that degrade both performance and efficiency. Recently, early-exit strategies are proposed to mitigate overthinking by dynamically and adaptively terminating redundant reasoning. However, current early-exit methods either introduce extra training overhead by relying on proxy models or limit inference throughput due to the frequent content switching between reasoning and generating probing answers. Moreover, most early-exit methods harm LRLMs performance due to over-truncation. Our insight stems from an observation: overthinking often causes LRLMs to deviate from the correct reasoning path, which is frequently accompanied by high-entropy transition tokens. Given this, we propose an early-exit method deeply coupled with the native reasoning process, which leverages the path deviation index as a dedicated monitoring metric for the frequent occurrence of high-entropy transition tokens to dynamically detect and terminate overthinking trajectories. We conduct experiments across multiple benchmarks using LRLMs of different types and scales, and the results indicate that our method delivers the largest performance improvement over vanilla CoT compared to existing early-exit methods.

Large language model (LLM) web agents are increasingly used for web navigation but remain far from human reliability on realistic, long-horizon tasks. Existing evaluations focus primarily on end-to-end success, offering limited insight into where failures arise. We propose a hierarchical planning framework to analyze web agents across three layers (i.e., high-level planning, low-level execution, and replanning), enabling process-based evaluation of reasoning, grounding, and recovery. Our experiments show that structured Planning Domain Definition Language (PDDL) plans produce more concise and goal-directed strategies than natural language (NL) plans, but low-level execution remains the dominant bottleneck. These results indicate that improving perceptual grounding and adaptive control, not only high-level reasoning, is critical for achieving human-level reliability. This hierarchical perspective provides a principled foundation for diagnosing and advancing LLM web agents.

Flow-matching policies hold great promise for reinforcement learning (RL) by capturing complex, multi-modal action distributions. However, their practical application is often hindered by prohibitive inference latency and ineffective online exploration. Although recent works have employed one-step distillation for fast inference, the structure of the initial noise distribution remains an overlooked factor that presents significant untapped potential. This overlooked factor, along with the challenge of controlling policy stochasticity, constitutes two critical areas for advancing distilled flow-matching policies. To overcome these limitations, we propose GoldenStart (GSFlow), a policy distillation method with Q-guided priors and explicit entropy control. Instead of initializing generation from uninformed noise, we introduce a Q-guided prior modeled by a conditional VAE. This state-conditioned prior repositions the starting points of the one-step generation process into high-Q regions, effectively providing a "golden start" that shortcuts the policy to promising actions. Furthermore, for effective online exploration, we enable our distilled actor to output a stochastic distribution instead of a deterministic point. This is governed by entropy regularization, allowing the policy to shift from pure exploitation to principled exploration. Our integrated framework demonstrates that by designing the generative startpoint and explicitly controlling policy entropy, it is possible to achieve efficient and exploratory policies, bridging the generative models and the practical actor-critic methods. We conduct extensive experiments on offline and online continuous control benchmarks, where our method significantly outperforms prior state-of-the-art approaches. Code will be available at https://github.com/ZhHe11/GSFlow-RL.

SystemVerilog Assertions (SVAs) are crucial for hardware verification. Recent studies leverage general-purpose LLMs to translate natural language properties to SVAs (NL2SVA), but they perform poorly due to limited data. We propose a data synthesis framework to tackle two challenges: the scarcity of high-quality real-world SVA corpora and the lack of reliable methods to determine NL-SVA semantic equivalence. For the former, large-scale open-source RTLs are used to guide LLMs to generate real-world SVAs; for the latter, bidirectional translation serves as a data selection method. With the synthesized data, we train CodeV-SVA, a series of SVA generation models. Notably, CodeV-SVA-14B achieves 75.8% on NL2SVA-Human and 84.0% on NL2SVA-Machine in Func.@1, matching or exceeding advanced LLMs like GPT-5 and DeepSeek-R1.

Federated learning (FL) allows distributed clients to collaboratively train a global model in a privacy-preserving manner. However, one major challenge is domain skew, where clients' data originating from diverse domains may hinder the aggregated global model from learning a consistent representation space, resulting in poor generalizable ability in multiple domains. In this paper, we argue that the domain skew is reflected in the domain-specific biased features of each client, causing the local model's representations to collapse into a narrow low-dimensional subspace. We then propose Federated Feature Decoupling and Calibration ($F^2$DC), which liberates valuable class-relevant information by calibrating the domain-specific biased features, enabling more consistent representations across domains. A novel component, Domain Feature Decoupler (DFD), is first introduced in $F^2$DC to determine the robustness of each feature unit, thereby separating the local features into domain-robust features and domain-related features. A Domain Feature Corrector (DFC) is further proposed to calibrate these domain-related features by explicitly linking discriminative signals, capturing additional class-relevant clues that complement the domain-robust features. Finally, a domain-aware aggregation of the local models is performed to promote consensus among clients. Empirical results on three popular multi-domain datasets demonstrate the effectiveness of the proposed $F^2$DC and the contributions of its two modules. Code is available at https://github.com/mala-lab/F2DC.

Designing correct UAV autonomy programs is challenging due to joint navigation, sensing and analytics requirements. While LLMs can generate code, their reliability for safety-critical UAVs remains uncertain. This paper presents AeroGen, an open-loop framework that enables consistently correct single-shot AI-generated drone control programs through structured guardrail prompting and integration with the AeroDaaS drone SDK. AeroGen encodes API descriptions, flight constraints and operational world rules directly into the system context prompt, enabling generic LLMs to produce constraint-aware code from user prompts, with minimal example code. We evaluate AeroGen across a diverse benchmark of 20 navigation tasks and 5 drone missions on urban, farm and inspection environments, using both imperative and declarative user prompts. AeroGen generates about 40 lines of AeroDaaS Python code in about 20s per mission, in both real-world and simulations, showing that structured prompting with a well-defined SDK improves robustness, correctness and deployability of LLM-generated drone autonomy programs.

Enterprises increasingly need natural language (NL) question answering over hybrid data lakes that combine structured tables and unstructured documents. Current deployed solutions, including RAG-based systems, typically rely on brute-force retrieval from each store and post-hoc merging. Such approaches are inefficient and leaky, and more critically, they lack explicit support for multi-hop reasoning, where a query is decomposed into successive steps (hops) that may traverse back and forth between structured and unstructured sources. We present Agentic DAG-Orchestrated Transformer (A.DOT) Planner, a framework for multi-modal, multi-hop question answering, that compiles user NL queries into directed acyclic graph (DAG) execution plans spanning both structured and unstructured stores. The system decomposes queries into parallelizable sub-queries, incorporates schema-aware reasoning, and applies both structural and semantic validation before execution. The execution engine adheres to the generated DAG plan to coordinate concurrent retrieval across heterogeneous sources, route intermediate outputs to dependent sub-queries, and merge final results in strict accordance with the plan's logical dependencies. Advanced caching mechanisms, incorporating paraphrase-aware template matching, enable the system to detect equivalent queries and reuse prior DAG execution plans for rapid re-execution, while the DataOps System addresses validation feedback or execution errors. The proposed framework not only improves accuracy and latency, but also produces explicit evidence trails, enabling verification of retrieved content, tracing of data lineage, and fostering user trust in the system's outputs. On benchmark dataset, A.DOT achieves 14.8% absolute gain in correctness and 10.7% in completeness over baselines.

Over-reliance on AI systems can undermine users' critical thinking and promote complacency, a risk intensified by the emergence of agentic AI systems that operate with minimal human involvement. In software engineering, agentic coding assistants are rapidly becoming embedded in everyday development workflows. Since software engineers create systems deployed across diverse and high-stakes real-world contexts, these assistants must function not merely as autonomous task performers but as Tools for Thought that actively support human reasoning and sensemaking. We conducted a formative study examining software engineers' cognitive engagement and sensemaking processes when working with an agentic coding assistant. Our findings reveal that cognitive engagement consistently declines as tasks progress, and that current agentic coding assistants' designs provide limited affordances for reflection, verification, and meaning-making. Based on these findings, e identify concrete design opportunities leveraging richer interaction modalities and cognitive-forcing mechanisms to sustain engagement and promote deeper thinking in AI-assisted programming.

The KV cache in self-attention has emerged as a major bottleneck in long-context and large-batch inference for LLMs. Existing approaches often treat sparsity prediction and compression as separate modules, relying on auxiliary index structures to select relevant tokens, and on complex quantization schemes to reduce memory usage. This fragmented design introduces redundant overhead and limits scalability. In this paper, we propose a novel paradigm: treating the compressed key representation not merely as storage, but as a self-indexing structure that directly enables efficient sparse attention. By designing a sign-based 1-bit vector quantization (VQ) scheme, our method unifies compression and retrieval in a single, hardware-friendly format. This approach eliminates the need for external indices or learning-based predictors, offering a lightweight yet robust solution for memory-constrained inference. All components are designed to be hardware-efficient and easy to implement. By implementing custom CUDA kernels, our method integrates seamlessly with FlashAttention, minimizing additional runtime and memory overhead. Experimental results demonstrate that our approach delivers both effectiveness and efficiency.

Contrastive pretraining models such as CLIP and CLAP underpin many vision-language and audio-language systems, yet their reliance on web-scale data raises growing concerns about memorizing Personally Identifiable Information (PII). Auditing such models via membership inference is challenging in practice: shadow-model MIAs are computationally prohibitive for large multimodal backbones, and existing multimodal attacks typically require querying the target with paired biometric inputs, thereby directly exposing sensitive biometric information to the target model. We propose Unimodal Membership Inference Detector (UMID), a text-only auditing framework that performs text-guided cross-modal latent inversion and extracts two complementary signals, similarity (alignment to the queried text) and variability (consistency across randomized inversions). UMID compares these statistics to a lightweight non-member reference constructed from synthetic gibberish and makes decisions via an ensemble of unsupervised anomaly detectors. Comprehensive experiments across diverse CLIP and CLAP architectures demonstrate that UMID significantly improves the effectiveness and efficiency over prior MIAs, delivering strong detection performance with sub-second auditing cost while complying with realistic privacy constraints.

Ethical decision governance has become a critical requirement for autonomous robotic systems operating in human-centered and safety-sensitive environments. This paper presents a real-time neuro-symbolic ethical governor designed to enable risk-aware supervisory control in autonomous robotic manipulation tasks. The proposed framework integrates transformer-based ethical reasoning with a probabilistic ethical risk field formulation and a threshold-based override control mechanism. language-grounded ethical intent inference capability is learned from natural language task descriptions using a fine-tuned DistilBERT model trained on the ETHICS commonsense dataset. A continuous ethical risk metric is subsequently derived from predicted unsafe action probability, confidence uncertainty, and probabilistic variance to support adaptive decision filtering. The effectiveness of the proposed approach is validated through simulated autonomous robot-arm task scenarios involving varying levels of human proximity and operational hazard. Experimental results demonstrate stable model convergence, reliable ethical risk discrimination, and improved safety-aware decision outcomes without significant degradation of task execution efficiency. The proposed neuro-symbolic architecture further provides enhanced interpretability compared with purely data-driven safety filters, enabling transparent ethical reasoning in real-time control loops. The findings suggest that ethical decision governance can be effectively modeled as a dynamic supervisory risk layer for autonomous robotic systems, with potential applicability to broader cyber-physical and assistive robotics domains.

We study the problem of evaluating the excess risk of large-scale empirical risk minimization under the square loss. Leveraging the idea of wild refitting and resampling, we assume only black-box access to the training algorithm and develop an efficient procedure for estimating the excess risk. Our evaluation algorithm is both computationally and data efficient. In particular, it requires access to only a single dataset and does not rely on any additional validation data. Computationally, it only requires refitting the model on several much smaller datasets obtained through sequential resampling, in contrast to previous wild refitting methods that require full-scale retraining and might therefore be unsuitable for large-scale trained predictors. Our algorithm has an interleaved sequential resampling-and-refitting structure. We first construct pseudo-responses through a randomized residual symmetrization procedure. At each round, we thus resample two sub-datasets from the resulting covariate pseudo-response pairs. Finally, we retrain the model separately on these two small artificial datasets. We establish high probability excess risk guarantees under both fixed design and random design settings, showing that with a suitably chosen noise scale, our interleaved resampling and refitting algorithm yields an upper bound on the prediction error. Our theoretical analysis draws on tools from empirical process theory, harmonic analysis, Toeplitz operator theory, and sharp tensor concentration inequalities.

In cognitive science and linguistic theory, dialogue is not seen as a chain of independent utterances but rather as a joint activity sustained by coherence, consistency, and shared understanding. However, many systems for open-domain and personalized dialogue use surface-level similarity metrics (e.g., BLEU, ROUGE, F1) as one of their main reporting measures, which fail to capture these deeper aspects of conversational quality. We re-examine a notable retrieval-augmented framework for personalized dialogue, LAPDOG, as a case study for evaluation methodology. Using both human and LLM-based judges, we identify limitations in current evaluation practices, including corrupted dialogue histories, contradictions between retrieved stories and persona, and incoherent response generation. Our results show that human and LLM judgments align closely but diverge from lexical similarity metrics, underscoring the need for cognitively grounded evaluation methods. Broadly, this work charts a path toward more reliable assessment frameworks for retrieval-augmented dialogue systems that better reflect the principles of natural human communication.

Image fusion aims to integrate complementary information from multiple source images to produce a more informative and visually consistent representation, benefiting both human perception and downstream vision tasks. Despite recent progress, most existing fusion methods are designed for specific tasks (i.e., multi-modal, multi-exposure, or multi-focus fusion) and struggle to effectively preserve source information during the fusion process. This limitation primarily arises from task-specific architectures and the degradation of source information caused by deep-layer propagation. To overcome these issues, we propose UniFusion, a unified image fusion framework designed to achieve cross-task generalization. First, leveraging DINOv3 for modality-consistent feature extraction, UniFusion establishes a shared semantic space for diverse inputs. Second, to preserve the understanding of each source image, we introduce a reconstruction-alignment loss to maintain consistency between fused outputs and inputs. Finally, we employ a bilevel optimization strategy to decouple and jointly optimize reconstruction and fusion objectives, effectively balancing their coupling relationship and ensuring smooth convergence. Extensive experiments across multiple fusion tasks demonstrate UniFusion's superior visual quality, generalization ability, and adaptability to real-world scenarios. Code is available at https://github.com/dusongcheng/UniFusion.

We proudly introduce Memory-as-Asset, a new memory paradigm towards human-centric artificial general intelligence (AGI). In this paper, we formally emphasize that human-centric, personal memory management is a prerequisite for complementing the collective knowledge of existing large language models (LLMs) and extending their knowledge boundaries through self-evolution. We introduce three key features that shape the Memory-as-Asset era: (1) Memory in Hand, which emphasizes human-centric ownership to maximize benefits to humans; (2) Memory Group, which provides collaborative knowledge formation to avoid memory islands, and (3) Collective Memory Evolution, which enables continuous knowledge growth to extend the boundary of knowledge towards AGI. We finally give a potential three-layer memory infrastructure to facilitate the Memory-as-Asset paradigm, with fast personal memory storage, an intelligent evolution layer, and a decentralized memory exchange network. Together, these components outline a foundational architecture in which personal memories become persistent digital assets that can be accumulated, shared, and evolved over time. We believe this paradigm provides a promising path toward scalable, human-centric AGI systems that continuously grow through the collective experiences of individuals and intelligent agents.

We present Vavanagi, a community-run platform for Hula (Vula'a), an Austronesian language of Papua New Guinea with approximately 10,000 speakers. Vavanagi supports crowdsourced English-Hula text translation and voice recording, with elder-led review and community-governed data infrastructure. To date, 77 translators and 4 reviewers have produced over 12k parallel sentence pairs covering 9k unique Hula words. We also propose a multi-level framework for measuring community involvement, from consultation to fully community-initiated and governed projects. We position Vavanagi at Level 5: initiative, design, implementation, and data governance all sit within the Hula community, making it, to our knowledge, the first community-led language technology initiative for a language of this size. Vavanagi shows how language technology can bridge village-based and urban members, connect generations, and support cultural heritage on the community's own terms.

A pictorial chart is an effective medium for visual storytelling, seamlessly integrating visual elements with data charts. However, creating such images is challenging because the flexibility of visual elements often conflicts with the rigidity of chart structures. This process thus requires a creative deformation that maintains both data faithfulness and visual aesthetics. Current methods that extract dense structural cues from natural images (e.g., edge or depth maps) are ill-suited as conditioning signals for pictorial chart generation. We present ChArtist, a domain-specific diffusion model for generating pictorial charts automatically, offering two distinct types of control: 1) spatial control that aligns well with the chart structure, and 2) subject-driven control that respects the visual characteristics of a reference image. To achieve this, we introduce a skeleton-based spatial control representation. This representation encodes only the data-encoding information of the chart, allowing for the easy incorporation of reference visuals without a rigid outline constraint. We implement our method based on the Diffusion Transformer (DiT) and leverage an adaptive position encoding mechanism to manage these two controls. We further introduce Spatially Gated Attention to modulate the interaction between spatial control and subject control. To support the fine-tuning of pre-trained models for this task, we created a large-scale dataset of 30,000 triplets (skeleton, reference image, pictorial chart). We also propose a unified data accuracy metric to evaluate the data faithfulness of the generated charts. We believe this work demonstrates that current generative models can achieve data-driven visual storytelling by moving beyond general-purpose conditions to task-specific representations. Project page: https://chartist-ai.github.io/.

Scene Text Image Super-Resolution (STISR) aims to restore high-resolution details in low-resolution text images, which is crucial for both human readability and machine recognition. Existing methods, however, often depend on external Optical Character Recognition (OCR) models for textual priors or rely on complex multi-component architectures that are difficult to train and reproduce. In this paper, we introduce DualTSR, a unified end-to-end framework that addresses both issues. DualTSR employs a single multimodal transformer backbone trained with a dual diffusion objective. It simultaneously models the continuous distribution of high-resolution images via Conditional Flow Matching and the discrete distribution of textual content via discrete diffusion. This shared design enables visual and textual information to interact at every layer, allowing the model to infer text priors internally instead of relying on an external OCR module. Compared with prior multi-branch diffusion systems, DualTSR offers a simpler end-to-end formulation with fewer hand-crafted components. Experiments on synthetic Chinese benchmarks and a curated real-world evaluation protocol show that DualTSR achieves strong perceptual quality and text fidelity.

Online market platforms play an increasingly powerful role in the economy. An empirical phenomenon is that platforms, such as Amazon, Apple, and DoorDash, also enter their own marketplaces, imitating successful products developed by third-party sellers. We formulate a Stackelberg model, where the platform acts as the leader by committing to an entry policy: when will it enter and compete on a product? We study this model through a theoretical and computational framework. We begin with a single seller, and consider different kinds of policies for entry. We characterize the seller's optimal explore-exploit strategy via a Gittins-index policy, and give an algorithm to compute the platform's optimal entry policy. We then consider multiple sellers, to account for competition and information spillover. Here, the Gittins-index characterization fails, and we employ deep reinforcement learning to examine seller equilibrium behavior. Our findings highlight the incentives that drive platform entry and seller innovation, consistent with empirical evidence from markets such as Amazon and Google Play, with implications for regulatory efforts to preserve innovation and market diversity.

Deploying trustworthy AI systems requires principled uncertainty quantification. Conformal prediction (CP) is a widely used framework for constructing prediction sets with distribution-free coverage guarantees. In many practical settings, including healthcare, finance, and mobile sensing, the calibration data required for CP are distributed across multiple clients, each with its own local data distribution. In this federated setting, data can often be partitioned into, potentially overlapping, groups, which may reflect client-specific strata or cross-cutting attributes such as demographic or semantic categories. We propose group-conditional federated conformal prediction (GC-FCP), a novel protocol that provides group-conditional coverage guarantees. GC-FCP constructs mergeable, group-stratified coresets from local calibration scores, enabling clients to communicate compact weighted summaries that support efficient aggregation and calibration at the server. Experiments on synthetic and real-world datasets validate the performance of GC-FCP compared to centralized calibration baselines.

YARA has established itself as the de facto standard for "Detection as Code," enabling analysts and DevSecOps practitioners to define signatures for malware identification across the software supply chain. Despite its pervasive use, the open-source YARA ecosystem remains characterized by ad-hoc sharing and opaque quality. Practitioners currently rely on public repositories without empirical evidence regarding the ecosystem's structural characteristics, maintenance and diffusion dynamics, or operational reliability. We conducted a large-scale mixed-method study of 8.4 million rules mined from 1,853 GitHub repositories. Our pipeline integrates repository mining to map supply chain dynamics, static analysis to assess syntactic quality, and dynamic benchmarking against 4,026 malware and 2,000 goodware samples to measure operational effectiveness. We reveal a highly centralized structure where 10 authors drive 80% of rule adoption. The ecosystem functions as a "static supply chain": repositories show a median inactivity of 782 days and a median technical lag of 4.2 years. While static quality scores appear high (mean = 99.4/100), operational benchmarking uncovers significant noise (false positives) and low recall. Furthermore, coverage is heavily biased toward legacy threats (Ransomware), leaving modern initial access vectors (Loaders, Stealers) severely underrepresented. These findings expose a systemic "double penalty": defenders incur high performance overhead for decayed intelligence. We argue that public repositories function as raw data dumps rather than curated feeds, necessitating a paradigm shift from ad-hoc collection to rigorous rule engineering. We release our dataset and pipeline to support future data-driven curation tools.

Gait recognition is an emerging biometric technology that enables non-intrusive and hard-to-spoof human identification. However, most existing methods are confined to short-range, unimodal settings and fail to generalize to long-range and cross-distance scenarios under real-world conditions. To address this gap, we present \textbf{LRGait}, the first LiDAR-Camera multimodal benchmark designed for robust long-range gait recognition across diverse outdoor distances and environments. We further propose \textbf{EMGaitNet}, an end-to-end framework tailored for long-range multimodal gait recognition. To bridge the modality gap between RGB images and point clouds, we introduce a semantic-guided fusion pipeline. A CLIP-based Semantic Mining (SeMi) module first extracts human body-part-aware semantic cues, which are then employed to align 2D and 3D features via a Semantic-Guided Alignment (SGA) module within a unified embedding space. A Symmetric Cross-Attention Fusion (SCAF) module hierarchically integrates visual contours and 3D geometric features, and a Spatio-Temporal (ST) module captures global gait dynamics. Extensive experiments on various gait datasets validate the effectiveness of our method.

Generative multimodal models can exhibit safety failures that are inherently relational: two benign concepts can become unsafe when linked by a specific action or relation (e.g., child-drinking-wine). Existing unlearning and concept-erasure approaches often target isolated concepts or image-text pairs, which can cause collateral damage to benign uses of the same objects and relations. We propose relationship-aware safety unlearning: a framework that explicitly represents unsafe object-relation-object (O-R-O) tuples and applies targeted parameter-efficient edits (LoRA) to suppress unsafe tuples while preserving object marginals and safe neighboring relations. We include CLIP-based experiments and robustness evaluation under paraphrase, contextual, and out-of-distribution image attacks.

Multimodal large language models (MLLMs) often suffer from perceptual impairments under extended reasoning modes, particularly in visual question answering (VQA) tasks. We identify attention dispersion as the underlying cause: during multi-step reasoning, the model's visual attention becomes scattered and drifts away from question-relevant regions, effectively "losing focus" on the visual input. To better understand this phenomenon, we analyze the attention maps of MLLMs and observe that reasoning prompts significantly reduce attention to regions critical for answering the question. We further find a strong correlation between the model's overall attention on image tokens and the spatial dispersiveness of its attention within the image. Leveraging this insight, we propose a training-free Visual Region-Guided Attention (VRGA) framework that selects visual heads based on an entropy-focus criterion and reweights their attention, effectively guiding the model to focus on question-relevant regions during reasoning. Extensive experiments on vision-language benchmarks demonstrate that our method effectively alleviates perceptual degradation, leading to improvements in visual grounding and reasoning accuracy while providing interpretable insights into how MLLMs process visual information.

The rapid expansion of electronic health record (EHR) systems has generated large volumes of unstructured clinical narratives that contain valuable information for disease identification, patient cohort discovery, and clinical decision support. Extracting structured knowledge from these free-text documents remains challenging because clinical language is highly specialized, labeled datasets are limited, and full fine-tuning of large pretrained language models can require substantial computational resources. Efficient adaptation strategies are therefore essential for practical clinical natural language processing applications. This study proposes a parameter-efficient selective fine-tuning framework for adapting GPT-2 to clinical text classification tasks. Instead of updating the entire pretrained model, the majority of network parameters are frozen, and only the final Transformer block, the final layer normalization module, and a lightweight classification head are updated during training. This design substantially reduces the number of trainable parameters while preserving the contextual representation capabilities learned during pretraining. The proposed approach is evaluated using radiology reports from the MIMIC-IV-Note dataset with automatically derived CheXpert-style labels. Experiments on 50,000 radiology reports demonstrate that selective fine-tuning achieves approximately 91% classification accuracy while updating fewer than 6% of the model parameters. Comparative experiments with head-only training and full-model fine-tuning show that the proposed method provides a favorable balance between predictive performance and computational efficiency. These results indicate that selective fine-tuning offers an efficient and scalable framework for clinical text classification.

Hyperkalemia is a life-threatening electrolyte disorder that is common in patients with chronic kidney disease and heart failure, yet frequent monitoring remains difficult outside hospital settings. We developed and validated Pocket-K, a single-lead AI-ECG system initialized from the ECGFounder foundation model for non-invasive hyperkalemia screening and handheld deployment. In this multicentre observational study using routinely collected clinical ECG and laboratory data, 34,439 patients contributed 62,290 ECG--potassium pairs. Lead I data were used to fine-tune the model. Data from Peking University People's Hospital were divided into development and temporal validation sets, and data from The Second Hospital of Tianjin Medical University served as an independent external validation set. Hyperkalemia was defined as venous serum potassium > 5.5 mmol/L. Pocket-K achieved AUROCs of 0.936 in internal testing, 0.858 in temporal validation, and 0.808 in external validation. For KDIGO-defined moderate-to-severe hyperkalemia (serum potassium >= 6.0 mmol/L), AUROCs increased to 0.940 and 0.861 in the temporal and external sets, respectively. External negative predictive value exceeded 99.3%. Model-predicted high risk below the hyperkalemia threshold was more common in patients with chronic kidney disease and heart failure. A handheld prototype enabled near-real-time inference, supporting future prospective evaluation in native handheld and wearable settings.

Multimodal Domain Generalization (MMDG) leverages the complementary strengths of multiple modalities to enhance model generalization on unseen domains. A central challenge in multimodal learning is optimization imbalance, where modalities converge at different speeds during training. This imbalance leads to unequal gradient contributions, allowing some modalities to dominate the learning process while others lag behind. Existing balancing strategies typically regulate each modality's gradient contribution based on its classification performance on the source domain to alleviate this issue. However, relying solely on source-domain accuracy neglects a key insight in MMDG: modalities that excel on the source domain may generalize poorly to unseen domains, limiting cross-domain gains. To overcome this limitation, we propose Gradient Modulation Projection (GMP), a unified strategy that promotes balanced optimization in MMDG. GMP first decouples gradients associated with classification and domain-invariance objectives. It then modulates each modality's gradient based on semantic and domain confidence. Moreover, GMP dynamically adjusts gradient projections by tracking the relative strength of each task, mitigating conflicts between classification and domain-invariant learning within modality-specific encoders. Extensive experiments demonstrate that GMP achieves state-of-the-art performance and integrates flexibly with diverse MMDG methods, significantly improving generalization across multiple benchmarks.

Anomaly detection for tabular data has been a long-standing unsupervised learning problem that remains a major challenge for current deep learning models. Recently, in-context learning has emerged as a new paradigm that has shifted efforts from task-specific optimization to large-scale pretraining aimed at creating foundation models that generalize across diverse datasets. Although in-context models, such as TabPFN, perform well in supervised problems, their learned classification-based priors may not readily extend to anomaly detection. In this paper, we study in-context models for anomaly detection and show that the unsupervised extensions to TabPFN exhibit unstable behavior, particularly in noisy or contaminated contexts, in addition to the high computational cost. We address these challenges and introduce TACTIC, an in-context anomaly detection approach based on pretraining with anomaly-centric synthetic priors, which provides fast and data-dependent reasoning about anomalies while avoiding dataset-specific tuning. In contrast to typical score-based approaches, which produce uncalibrated anomaly scores that require post-processing (e.g. threshold selection or ranking heuristics), the proposed model is trained as a discriminative predictor, enabling unambiguous anomaly decisions in a single forward pass. Through experiments on real-world datasets, we examine the performance of TACTIC in clean and noisy contexts with varying anomaly rates and different anomaly types, as well as the impact of prior choices on detection quality. Our experiments clearly show that specialized anomaly-centric in-context models such as TACTIC are highly competitive compared to other task-specific methods.

Average treatment effects (ATE) and conditional average treatment effects (CATE) are foundational causal estimands, but they target changes in expected outcomes and can miss treatment-induced changes in the shape of outcome distributions. A canonical failure mode occurs when control outcomes are unimodal, treated outcomes become bimodal, and both distributions have the same mean. In such cases mean-based causal estimands are zero even though the geometry and topology of the outcome law change substantially. This paper develops a topological causal framework based on persistent homology. We formalize a persistent-homology ignorability condition, define topological analogues of CATE and ATE, and prove that these estimands are identifiable up to an explicit error bound under approximate topological ignorability. We also clarify a subtle but important point: a marginal persistence-diagram effect is not identified from conditional topological ignorability alone because persistent homology does not in general commute with mixtures over covariates. To preserve the original intuition while ensuring scientific correctness, we retain the marginal effect as a motivating quantity, but place the mathematically sound conditional estimands at the center of the theory. A synthetic experiment with mean-preserving topology change shows that mean-based causal estimands remain near zero while the proposed topological effect increases sharply and remains recoverable after adjustment for confounding.

Many disciplines need quantitative models that synthesize experimental data across multiple instances of the same general system. For example, neuroscientists must combine data from the brains of many individual animals to understand the species' brain in general. However, typical machine learning models treat one system instance at a time. Here we introduce a machine learning framework, deep probabilistic model synthesis (DPMS), that leverages system properties auxiliary to the model to combine data across system instances. DPMS specifically uses variational inference to learn a conditional prior distribution and instance-specific posterior distributions over model parameters that respectively tie together the system instances and capture their unique structure. DPMS can synthesize a wide variety of model classes, such as those for regression, classification, and dimensionality reduction, and we demonstrate its ability to improve upon single-instance models on synthetic data and whole-brain neural activity data from larval zebrafish.

Large language models (LLMs) are entering clinician workflows, yet evaluations rarely measure how clinician reasoning shapes model behavior during clinical interactions. We combined 61 New England Journal of Medicine Case Records with 92 real-world clinician-AI interactions to evaluate 21 reasoning LLM variants across 8 frontier models on differential diagnosis generation and next step recommendations under three conditions: reasoning alone, after expert clinician context, and after adversarial clinician context. LLM-clinician concordance increased substantially after clinician exposure, with simulations sharing >=3 differential diagnosis items rising from 65.8% to 93.5% and >=3 next step recommendations from 20.3% to 53.8%. Expert context significantly improved correct final diagnosis inclusion across all 21 models (mean +20.4 percentage points), reflecting both reasoning improvement and passive content echoing, while adversarial context caused significant diagnostic degradation in 14 models (mean -5.4 percentage points). Multi-turn disagreement probes revealed distinct model phenotypes ranging from highly conformist to dogmatic, with adversarial arguments remaining a persistent vulnerability even for otherwise resilient models. Inference-time scaling reduced harmful echoing of clinician-introduced recommendations across WHO-defined harm severity tiers (relative reductions: 62.7% mild, 57.9% moderate, 76.3% severe, 83.5% death-tier). In GPT-4o experiments, explicit clinician uncertainty signals improved diagnostic performance after adversarial context (final diagnosis inclusion 27% to 42%) and reduced alignment with incorrect arguments by 21%. These findings establish a foundation for evaluating clinician-AI collaboration, introducing interactive metrics and mitigation strategies essential for safety and robustness.

In neural network models, soft mixtures of fixed candidate components (e.g., logic gates and sub-networks) are often used during training for stable optimization, while hard selection is typically used at inference. This raises questions about training-inference mismatch. We analyze this gap by separating forward-pass computation (hard selection vs. soft mixture) from stochasticity (with vs. without Gumbel noise). Using logic gate networks as a testbed, we observe distinct behaviors across four methods: Hard-ST achieves zero selection gap by construction; Gumbel-ST achieves near-zero gap when training succeeds but suffers accuracy collapse at low temperatures; Soft-Mix achieves small gap only at low temperature via weight concentration; and Soft-Gumbel exhibits large gaps despite Gumbel noise, confirming that noise alone does not reduce the gap. We propose CAGE (Confidence-Adaptive Gradient Estimation) to maintain gradient flow while preserving forward alignment. On logic gate networks, Hard-ST with CAGE achieves over 98% accuracy on MNIST and over 58% on CIFAR-10, both with zero selection gap across all temperatures, while Gumbel-ST without CAGE suffers a 47-point accuracy collapse.

The generative artificial intelligence (AI) ecosystem is undergoing rapid transformations that threaten its sustainability. As models transition from research prototypes to high-traffic products, the energetic burden has shifted from one-time training to recurring, unbounded inference. This is exacerbated by reasoning models that inflate compute costs by orders of magnitude per query. The prevailing pursuit of artificial general intelligence through scaling of monolithic models is colliding with hard physical constraints: grid failures, water consumption, and diminishing returns on data scaling. This trajectory yields models with impressive factual recall but struggles in domains requiring in-depth reasoning, possibly due to insufficient abstractions in training data. Current large language models (LLMs) exhibit genuine reasoning depth only in domains like mathematics and coding, where rigorous, pre-existing abstractions provide structural grounding. In other fields, the current approach fails to generalize well. We propose an alternative trajectory based on domain-specific superintelligence (DSS). We argue for first constructing explicit symbolic abstractions (knowledge graphs, ontologies, and formal logic) to underpin synthetic curricula enabling small language models to master domain-specific reasoning without the model collapse problem typical of LLM-based synthetic data methods. Rather than a single generalist giant model, we envision "societies of DSS models": dynamic ecosystems where orchestration agents route tasks to distinct DSS back-ends. This paradigm shift decouples capability from size, enabling intelligence to migrate from energy-intensive data centers to secure, on-device experts. By aligning algorithmic progress with physical constraints, DSS societies move generative AI from an environmental liability to a sustainable force for economic empowerment.

Multimodal Large Language Models (MLLMs) have shown strong performance in visual and audio understanding when evaluated in isolation. However, their ability to jointly reason over omni-modal (visual, audio, and textual) signals in long and complex videos remains largely unexplored. We introduce MMOU, a new benchmark designed to systematically evaluate multimodal understanding and reasoning under these challenging, real-world conditions. MMOU consists of 15,000 carefully curated questions paired with 9038 web-collected videos of varying length, spanning diverse domains and exhibiting rich, tightly coupled audio-visual content. The benchmark covers 13 fundamental skill categories, all of which require integrating evidence across modalities and time. All questions are manually annotated across multiple turns by professional annotators, ensuring high quality and reasoning fidelity. We evaluate 20+ state-of-the-art open-source and proprietary multimodal models on MMOU. The results expose substantial performance gaps: the best closed-source model achieves only 64.2% accuracy, while the strongest open-source model reaches just 46.8%. Our results highlight the challenges of long-form omni-modal understanding, revealing that current models frequently fail to apply even fundamental skills in long videos. Through detailed analysis, we further identify systematic failure modes and provide insights into where and why current models break.

High-fidelity computational fluid dynamics (CFD) simulations are widely used to analyze nuclear reactor transients, but are computationally expensive when exploring large parameter spaces. Multifidelity surrogate models offer an approach to reduce cost by combining information from simulations of varying resolution. In this work, several multifidelity machine learning methods were evaluated for predicting the time to onset of natural circulation (ONC) and the temperature after ONC for a high-temperature gas reactor (HTGR) depressurized loss of forced cooling transient. A CFD model was developed in Ansys Fluent to generate 1000 simulation samples at each fidelity level, with low and medium-fidelity datasets produced by systematically coarsening the high-fidelity mesh. Multiple surrogate approaches were investigated, including multifidelity Gaussian processes and several neural network architectures, and validated on analytical benchmark functions before application to the ONC dataset. The results show that performance depends strongly on the informativeness of the input variables and the relationship between fidelity levels. Models trained using dominant inputs identified through prior sensitivity analysis consistently outperformed models trained on the full input set. The low- and high-fidelity pairing produced stronger performance than configurations involving medium-fidelity data, and two-fidelity configurations generally matched or exceeded three-fidelity counterparts at equivalent computational cost. Among the methods evaluated, multifidelity GP provided the most robust performance across input configurations, achieving excellent metrics for both time to ONC and temperature after ONC, while neural network approaches achieved comparable accuracy with substantially lower training times.

Correlated equilibria enable a coordinator to influence the self-interested agents by recommending actions that no player has an incentive to deviate from. However, the effectiveness of this mechanism relies on accurate knowledge of the agents' cost structures. When cost parameters are uncertain, the recommended actions may no longer be incentive compatible, allowing agents to benefit from deviating from them. We study a chance-constrained correlated equilibrium problem formulation that accounts for uncertainty in agents' costs and guarantees incentive compatibility with a prescribed confidence level. We derive sensitivity results that quantify how uncertainty in individual incentive constraints affects the expected coordination outcome. In particular, the analysis characterizes the value of information by relating the marginal benefit of reducing uncertainty to the dual sensitivities of the incentive constraints, providing guidance on which sources of uncertainty should be prioritized for information acquisition. The results further reveal that increasing the confidence level is not always beneficial and can introduce a tradeoff between robustness and system efficiency. Numerical experiments demonstrate that the proposed framework maintains coordination performance in uncertain environments and are consistent with the theoretical insights developed in the analysis.

This study presents a conditional flow matching framework for solving physics-constrained Bayesian inverse problems. In this setting, samples from the joint distribution of inferred variables and measurements are assumed available, while explicit evaluation of the prior and likelihood densities is not required. We derive a simple and self-contained formulation of both the unconditional and conditional flow matching algorithms, tailored specifically to inverse problems. In the conditional setting, a neural network is trained to learn the velocity field of a probability flow ordinary differential equation that transports samples from a chosen source distribution directly to the posterior distribution conditioned on observed measurements. This black-box formulation accommodates nonlinear, high-dimensional, and potentially non-differentiable forward models without restrictive assumptions on the noise model. We further analyze the behavior of the learned velocity field in the regime of finite training data. Under mild architectural assumptions, we show that overtraining can induce degenerate behavior in the generated conditional distributions, including variance collapse and a phenomenon termed selective memorization, wherein generated samples concentrate around training data points associated with similar observations. A simplified theoretical analysis explains this behavior, and numerical experiments confirm it in practice. We demonstrate that standard early-stopping criteria based on monitoring test loss effectively mitigate such degeneracy. The proposed method is evaluated on several physics-based inverse problems. We investigate the impact of different choices of source distributions, including Gaussian and data-informed priors. Across these examples, conditional flow matching accurately captures complex, multimodal posterior distributions while maintaining computational efficiency.

Many software development platforms now support LLM-driven programming, or "vibe coding", a technique that allows one to specify programs in natural language and iterate from observed behavior, all without directly editing source code. While its adoption is accelerating, little is known about which skills best predict success in this workflow. We report a preregistered cross-sectional study with tertiary-level students (N = 100) who completed measures of computer-science achievement, domain-general cognitive skills, written-communication proficiency, and a vibe-coding assessment. Tasks were curated via an eight-expert consensus process and executed in a purpose-built, vibe-coding environment that mirrors commercial tools while enabling controlled evaluation. We find that both writing skill and CS achievement are significant predictors of vibe-coding performance, and that CS achievement remains a significant predictor after controlling for domain-general cognitive skills. The results may inform tool and curriculum design, including when to emphasize prompt-writing versus CS fundamentals to support future software creators.

Automated segmentation of Martian landslides, particularly in tectonically active regions such as Valles Marineris,is important for planetary geology, hazard assessment, and future robotic exploration. However, detecting landslides from planetary imagery is challenging due to the heterogeneous nature of available sensing modalities and the limited number of labeled samples. Each observation combines RGB imagery with geophysical measurements such as digital elevation models, slope maps, thermal inertia, and contextual grayscale imagery, which differ significantly in resolution and statistical properties. To address these challenges, we propose DualSwinFusionSeg, a multimodal segmentation architecture that separates modality-specific feature extraction and performs multi-scale cross-modal fusion. The model employs two parallel Swin Transformer V2 encoders to independently process RGB and auxiliary geophysical inputs, producing hierarchical feature representations. Corresponding features from the two streams are fused at multiple scales and decoded using a UNet++ decoder with dense nested skip connections to preserve fine boundary details. Extensive ablation studies evaluate modality contributions, loss functions, decoder architectures, and fusion strategies. Experiments on the MMLSv2 dataset from the PBVS 2026 Mars-LS Challenge show that modality-specific encoders and simple concatenation-based fusion improve segmentation accuracy under limited training data. The final model achieves 0.867 mIoU and 0.905 F1 on the development benchmark and 0.783 mIoU on the held-out test set, demonstrating strong performance for multimodal planetary surface segmentation.

We evaluate JEPA-style predictive representation learning versus reconstruction-based autoencoders on a controlled "TV-series" linear dynamical system with known latent state and a single noise parameter. While an initial comparison suggests JEPA is markedly more robust to noise, further diagnostics show that autoencoder failures are strongly influenced by asymmetries in objectives and by bottleneck/component-selection effects (confirmed by PCA baselines). Motivated by these findings, we introduce gated predictive autoencoders that learn to select predictable components, mimicking the beneficial feature-selection behavior observed in over-parameterized PCA. On this toy testbed, the proposed gated model is stable across noise levels and matches or outperforms JEPA.

This paper introduces GELATO (Government, Executive, Legislative, and Treaty Ontology), a dataset of U.S. House and Senate bills from the 118th Congress annotated using a novel two-level named entity recognition ontology designed for U.S. legislative texts. We fine-tune transformer-based models (BERT, RoBERTa) of different architectures and sizes on this dataset for first-level prediction. We then use LLMs with optimized prompts to complete the second level prediction. The strong performance of RoBERTa and relatively weak performance of BERT models, as well as the application of LLMs as second-level predictors, support future research in legislative NER or downstream tasks using these model combinations as extraction tools.

Diffusion and flow models achieve State-Of-The-Art (SOTA) generative performance, yet many practically important behaviors such as fine-grained prompt fidelity, compositional correctness, and text rendering are weakly specified by score or flow matching pretraining objectives. Reinforcement Learning (RL) fine-tuning with external, black-box rewards is a natural remedy, but diffusion RL is often brittle. Trajectory-based methods incur high memory cost and high-variance gradient estimates; forward-process approaches converge faster but can suffer from distribution drift, and hence reward hacking. In this work, we present \textbf{Centered Reward Distillation (CRD)}, a diffusion RL framework derived from KL-regularized reward maximization built on forward-process-based fine-tuning. The key insight is that the intractable normalizing constant cancels under \emph{within-prompt centering}, yielding a well-posed reward-matching objective. To enable reliable text-to-image fine-tuning, we introduce techniques that explicitly control distribution drift: (\textit{i}) decoupling the sampler from the moving reference to prevent ratio-signal collapse, (\textit{ii}) KL anchoring to a CFG-guided pretrained model to control long-run drift and align with the inference-time semantics of the pre-trained model, and (\textit{iii}) reward-adaptive KL strength to accelerate early learning under large KL regularization while reducing late-stage exploitation of reward-model loopholes. Experiments on text-to-image post-training with \texttt{GenEval} and \texttt{OCR} rewards show that CRD achieves competitive SOTA reward optimization results with fast convergence and reduced reward hacking, as validated on unseen preference metrics.

Classical scaling laws model AI performance as monotonically improving with model size. We challenge this assumption by deriving the Institutional Scaling Law, showing that institutional fitness -- jointly measuring capability, trust, affordability, and sovereignty -- is non-monotonic in model scale, with an environment-dependent optimum N*(epsilon). Our framework extends the Sustainability Index of Han et al. (2025) from hardware-level to ecosystem-level analysis, proving that capability and trust formally diverge beyond critical scale (Capability-Trust Divergence). We further derive a Symbiogenetic Scaling correction demonstrating that orchestrated systems of domain-specific models can outperform frontier generalists in their native deployment environments. These results are contextualized within a formal evolutionary taxonomy of generative AI spanning five eras (1943-present), with analysis of frontier lab dynamics, sovereign AI emergence, and post-training alignment evolution from RLHF through GRPO. The Institutional Scaling Law predicts that the next phase transition will be driven not by larger models but by better-orchestrated systems of domain-specific models adapted to specific institutional niches.

Honeypots are deception systems that emulate vulnerable services to collect threat intelligence. While deploying many honeypots increases the opportunity to observe attacker behaviour, in practise network and computational resources limit the number of honeypots that can be exposed. Hence, practitioners must select the assets to deploy, a decision that is typically made statically despite attackers' tactics evolving over time. This work investigates an AI-driven agentic architecture that autonomously manages honeypot exposure in response to ongoing attacks. The proposed agent analyses Intrusion Detection System (IDS) alerts and network state to infer the progression of the attack, identify compromised assets, and predict likely attacker targets. Based on this assessment, the agent dynamically reconfigures the system to maintain attacker engagement while minimizing unnecessary exposure. The approach is evaluated in a simulated environment where attackers execute Proof-of-Concept exploits for known CVEs. Preliminary results indicate that the agent can effectively infer the intent of the attacker and improve the efficiency of exposure under resource constraints

Sentence simplification aims to make complex text more accessible by reducing linguistic complexity while preserving the original meaning. However, progress in this area remains limited for mid-resource and low-resource languages due to the scarcity of high-quality data. To address this gap, we introduce the OasisSimp dataset, a multilingual dataset for sentence-level simplification covering five languages: English, Sinhala, Tamil, Pashto, and Thai. Among these, no prior sentence simplification datasets exist for Thai, Pashto, and Tamil, while limited data is available for Sinhala. Each language simplification dataset was created by trained annotators who followed detailed guidelines to simplify sentences while maintaining meaning, fluency, and grammatical correctness. We evaluate eight open-weight multilingual Large Language Models (LLMs) on the OasisSimp dataset and observe substantial performance disparities between high-resource and low-resource languages, highlighting the simplification challenges in multilingual settings. The OasisSimp dataset thus provides both a valuable multilingual resource and a challenging benchmark, revealing the limitations of current LLM-based simplification methods and paving the way for future research in low-resource sentence simplification. The dataset is available at https://OasisSimpDataset.github.io/.

Contextual sparsity is one of the approaches used to reduce computational complexity in the inference process of large language models (LLMs). Existing techniques for efficient LLM inference acceleration based on contextual sparsity with minimal accuracy degradation require training sparse pattern predictors. This paper presents a framework for accelerating inference of ReGLU-based feed-forward networks (FFNs) within LLMs. The proposed framework provides a fast, training-free method for building sparse pattern predictors using truncation-aware singular value decomposition (SVD) of the gate projection matrix, along with a threshold calibration algorithm, and inference executors supporting conditional computation on CUDA and CANN devices. Experiments on three sparse LLMs with an average activation sparsity level of 90% in the FFNs demonstrate up to a 1.8x reduction in end-to-end decoding time while maintaining less than 1% degradation in benchmark scores on tasks involving complex math and code generation. This work advances the deployment of LLMs on edge devices.

Climate-vulnerable road networks require a paradigm shift from reactive, fix-on-failure repairs to predictive, decision-ready maintenance. This paper introduces ST-ResGAT, a novel Spatio-Temporal Residual Graph Attention Network that fuses residual graph-attention encoding with GRU temporal aggregation to forecast pavement deterioration. Engineered for resource-constrained deployment, the framework translates continuous Pavement Condition Index (PCI) forecasts directly into the American Society for Testing and Materials (ASTM)-compliant maintenance priorities. Using a real-world inspection dataset of 750 segments in Sylhet, Bangladesh (2021-2024), ST-ResGAT significantly outperforms traditional non-spatial machine learning baselines, achieving exceptional predictive fidelity (R2 = 0.93, RMSE = 2.72). Crucially, ablation testing confirmed the mathematical necessity of modeling topological neighbor effects, proving that structural decay acts as a spatial contagion. Uniquely, we integrate GNNExplainer to unbox the model, demonstrating that its learned priorities align perfectly with established physical engineering theory. Furthermore, we quantify classification safety: achieving 85.5% exact ASTM class agreement and 100% adjacent-class containment, ensuring bounded, engineer-safe predictions. To connect model outputs to policy, we generate localized longitudinal maintenance profiles, perform climate stress-testing, and derive Pareto sustainability frontiers. ST-ResGAT therefore offers a practical, explainable, and sustainable blueprint for intelligent infrastructure management in high-risk, low-resource geological settings.

In recent years, machine learning (ML) based software systems are increasingly deployed in several critical applications, yet systematic testing of their behavior remains challenging due to complex model architectures, large input spaces, and evolving deployment environments. Existing testing approaches often rely on generating test cases based on given requirements, which often fail to reveal critical bugs of modern ML models due to their complex nature. Most importantly, such approaches, although they can be used to detect the presence of specific failures in the ML software, they hardly provide any message as to how to fix such errors. To tackle this, in this paper, we present DeepFix, a tool for automated testing of the entire ML pipeline using an agentic AI framework. Our testing approach first leverages Deepchecks to test the ML software for any potential bugs, and thereafter, uses an agentic AI-based approach to generate a detailed bug report. This includes a ranking, based on the severity of the found bugs, along with their explanations, which can be interpreted easily by any non-data science experts and most importantly, also provides possible ways to fix these bugs. Additionally, DeepFix supports several types of ML software systems and can be integrated easily to any ML workflow, enabling continuous testing throughout the development lifecycle. We discuss our already validated cases as well as some planned validations designed to demonstrate how the agentic testing process can reveal hidden failure modes that remain undetected by conventional testing methods. A 5-minute screencast demonstrating the tool's core functionality is available at https://youtu.be/WfwZmFcQgBQ.

Trustworthiness in artificial intelligence depends not only on what a model decides, but also on how it handles and explains cases in which a reliable decision cannot be made. In critical domains such as healthcare and finance, a reject option allows the model to abstain when evidence is insufficient, making it essential to explain why an instance is rejected in order to support informed human intervention. In these settings, explanations must not only be interpretable, but also faithful to the underlying model and computationally efficient enough to support real-time decision making. Abductive explanations guarantee fidelity, but their exact computation is known to be NP-hard for many classes of models, limiting their practical applicability. Computing \textbf{minimum-size} abductive explanations is an even more challenging problem, as it requires reasoning not only about fidelity but also about optimality. Prior work has addressed this challenge in restricted settings, including log-linear-time algorithms for computing minimum-size abductive explanations in linear models without rejection, as well as a polynomial-time method based on linear programming for computing abductive explanations, without guarantees of minimum size, for linear models with a reject option. In this work, we bridge these lines of research by computing minimum-size abductive explanations for linear models with a reject option. For accepted instances, we adapt the log-linear algorithm to efficiently compute optimal explanations. For rejected instances, we formulate a 0-1 integer linear programming problem that characterizes minimum-size abductive explanations of rejection. Although this formulation is NP-hard in theory, our experimental results show that it is consistently more efficient in practice than the linear-programming-based approach that does not guarantee minimum-size explanations.

We address the brittleness of Bayesian experimental design under model misspecification by formulating the problem as a max--min game between the experimenter and an adversarial nature subject to information-theoretic constraints. We demonstrate that this approach yields a robust objective governed by Sibson's $α$-mutual information~(MI), which identifies the $α$-tilted posterior as the robust belief update and establishes the Rényi divergence as the appropriate measure of conditional information gain. To mitigate the bias and variance of nested Monte Carlo estimators needed to estimate Sibson's $α$-MI, we adopt a PAC-Bayes framework to search over stochastic design policies, yielding rigorous high-probability lower bounds on the robust expected information gain that explicitly control finite-sample error.

As modern text-to-image (T2I) models draw closer to synthesizing highly realistic content, the threat of unsafe content generation grows, and it becomes paramount to exercise control. Existing approaches steer these models by applying Euclidean adjustments to text embeddings, redirecting the generation away from unsafe concepts. In this work, we introduce hyperbolic control (HyCon): a novel control mechanism based on parallel transport that leverages semantically aligned hyperbolic representation space to yield more expressive and stable manipulation of concepts. HyCon reuses off-the-shelf generative models and a state-of-the-art hyperbolic text encoder, linked via a lightweight adapter. HyCon achieves state-of-the-art results across four safety benchmarks and four T2I backbones, showing that hyperbolic steering is a practical and flexible approach for more reliable T2I generation.

The Expected Calibration Error (ece), the dominant calibration metric in machine learning, compares predicted probabilities against empirical frequencies of binary outcomes. This is appropriate when labels are binary events. However, many modern settings produce labels that are themselves probabilities rather than binary outcomes: a radiologist's stated confidence, a teacher model's soft output in knowledge distillation, a class posterior derived from a generative model, or an annotator agreement fraction. In these settings, ece commits a category error - it discards the probabilistic information in the label by forcing it into a binary comparison. The result is not a noisy approximation that more data will correct. It is a structural misalignment that persists and converges to the wrong answer with increasing precision as sample size grows. We introduce the Soft Mean Expected Calibration Error (smece), a calibration metric for settings where labels are of probabilistic nature. The modification to the ece formula is one line: replace the empirical hard-label fraction in each prediction bin with the mean probability label of the samples in that bin. smece reduces exactly to ece when labels are binary, making it a strict generalisation.

Space missions increasingly deploy high-fidelity sensors that produce data volumes exceeding onboard buffering and downlink capacity. This work evaluates FPGA acceleration of neural networks (NNs) across four space use cases on the AMD ZCU104 board. We use Vitis AI (AMD DPU) and Vitis HLS to implement inference, quantify throughput and energy, and expose toolchain and architectural constraints relevant to deployment. Vitis AI achieves up to 34.16$\times$ higher inference rate than the embedded ARM CPU baseline, while custom HLS designs reach up to 5.4$\times$ speedup and add support for operators (e.g., sigmoids, 3D layers) absent in the DPU. For these implementations, measured MPSoC inference power spans 1.5-6.75 W, reducing energy per inference versus CPU execution in all use cases. These results show that NN FPGA acceleration can enable onboard filtering, compression, and event detection, easing downlink pressure in future missions.

Trained Transformers have been shown to compute abstract features that appear redundant for predicting the immediate next token. We identify which components of the gradient signal from the next-token prediction objective give rise to this phenomenon, and we propose a method to estimate the influence of those components on the emergence of specific features. After validating our approach on toy tasks, we use it to interpret the origins of the world model in OthelloGPT and syntactic features in a small language model. Finally, we apply our framework to a pretrained LLM, showing that features with extremely high or low influence on future tokens tend to be related to formal reasoning domains such as code. Overall, our work takes a step toward understanding hidden features of Transformers through the lens of their development during training.

Estimating multi-component T2 relaxation distributions from Multi-Echo Spin Echo (MESE) MRI is a severely ill-posed inverse problem, traditionally solved using regularized non-negative least squares (NNLS). In abdominal imaging, particularly the pancreas, low SNR and residual uncorrelated noise challenge classical solvers and deterministic deep learning models. We introduce a bootstrap-based inference framework for robust distributional T2 estimation that performs stochastic resampling of the echo train and aggregates predictions across multiple subsets. This treats the acquisition as a distribution rather than a fixed input, yielding variance-reduced, physically consistent estimates and converting deterministic relaxometry networks into probabilistic ensemble predictors. Applied to the P2T2 architecture, our method uses inference-time bootstrapping to smooth noise artifacts and enhance fidelity to the underlying relaxation distribution. Noninvasive pancreatic evaluation is limited by location and biopsy risks, highlighting the need for biomarkers capable of capturing early pathophysiological changes. In type 1 diabetes (T1DM), progressive beta-cell destruction begins years before overt hyperglycemia, yet current imaging cannot assess early islet decline. We evaluate clinical utility via a test-retest reproducibility study (N=7) and a T1DM versus healthy differentiation task (N=8). Our approach achieves the lowest Wasserstein distances across repeated scans and superior sensitivity to physiology-driven shifts in the relaxation-time distribution, outperforming NNLS and deterministic deep learning baselines. These results establish inference-time bootstrapping as an effective enhancement for quantitative T2 relaxometry in low-SNR abdominal imaging.

Textual Emotion Classification (TEC) is one of the most difficult NLP tasks. State of the art approaches rely on Large language models (LLMs) and multi-model ensembles. In this study, we challenge the assumption that larger scale or more complex models are necessary for improved performance. In order to improve logical consistency, We introduce CMHL, a novel single-model architecture that explicitly models the logical structure of emotions through three key innovations: (1) multi-task learning that jointly predicts primary emotions, valence, and intensity, (2) psychologically-grounded auxiliary supervision derived from Russell's circumplex model, and (3) a novel contrastive contradiction loss that enforces emotional consistency by penalizing mutually incompatible predictions (e.g., simultaneous high confidence in joy and anger). With just 125M parameters, our model outperforms 56x larger LLMs and sLM ensembles with a new state-of-the-art F1 score of 93.75\% compared to (86.13\%-93.2\%) on the dair-ai Emotion dataset. We further show cross domain generalization on the Reddit Suicide Watch and Mental Health Collection dataset (SWMH), outperforming domain-specific models like MentalBERT and MentalRoBERTa with an F1 score of 72.50\% compared to (68.16\%-72.16\%) + a 73.30\% recall compared to (67.05\%-70.89\%) that translates to enhanced sensitivity for detecting mental health distress. Our work establishes that architectural intelligence (not parameter count) drives progress in TEC. By embedding psychological priors and explicit consistency constraints, a well-designed single model can outperform both massive LLMs and complex ensembles, offering a efficient, interpretable, and clinically-relevant paradigm for affective computing.

The classification of mental health is challenging for a variety of reasons. For one, there is overlap between the mental health issues. In addition, the signs of mental health issues depend on the context of the situation, making classification difficult. Although fine-tuning transformers has improved the performance for mental health classification, standard cross-entropy training tends to create entangled feature spaces and fails to utilize all the information the transformers contain. We present a new framework that focuses on representations to improve mental health classification. This is done using two methods. First, \textbf{layer-attentive residual aggregation} which works on residual connections to to weigh and fuse representations from all transformer layers while maintaining high-level semantics. Second, \textbf{supervised contrastive feature learning} uses temperature-scaled supervised contrastive learning with progressive weighting to increase the geometric margin between confusable mental health problems and decrease class overlap by restructuring the feature space. With a score of \textbf{74.36\%}, the proposed method is the best performing on the SWMH benchmark and outperforms models that are domain-specialized, such as \textit{MentalBERT} and \textit{MentalRoBERTa} by margins of (3.25\% - 2.2\%) and 2.41 recall points over the highest achieving model. These findings show that domain-adaptive pretraining for mental health text classification can be surpassed by carefully designed representation geometry and layer-aware residual integration, which also provide enhanced interpretability through learnt layer importance.

Super-resolution (SR) of satellite imagery is challenging due to the lack of paired low-/high-resolution data. Recent self-supervised SR methods overcome this limitation by exploiting the temporal redundancy in burst observations, but they lack a mechanism to quantify uncertainty in the reconstruction. In this work, we introduce a novel self-supervised loss that allows to estimate uncertainty in image super-resolution without ever accessing the ground-truth high-resolution data. We adopt a decision-theoretic perspective and show that minimizing the corresponding Bayesian risk yields the posterior mean and variance as optimal estimators. We validate our approach on a synthetic SkySat L1B dataset and demonstrate that it produces calibrated uncertainty estimates comparable to supervised methods. Our work bridges self-supervised restoration with uncertainty quantification, making a practical framework for uncertainty-aware image reconstruction.

Despite recent advances in Text-to-Video (T2V) synthesis, generating high-fidelity and dynamic motion remains a significant challenge. Existing methods primarily rely on Classifier-Free Guidance (CFG), often with explicit negative prompts (e.g. "static", "blurry"), to suppress undesired artifacts. However, such explicit negations frequently introduce unintended semantic bias and distort object integrity; a phenomenon we define as Content-Motion Drift. To address this, we propose MotionCFG, a framework that enhances motion dynamics by contrasting a target concept with its noise-perturbed counterparts. Specifically, by injecting Gaussian noise into the concept embeddings, MotionCFG creates localized negative anchors that encapsulate a broad complementary space of sub-optimal motion variations. Unlike explicit negations, this approach facilitates implicit hard negative mining without shifting the global semantic identity, allowing for a focused refinement of temporal details. Combined with a piecewise guidance schedule that confines intervention to the early denoising steps, MotionCFG consistently improves motion dynamics across state-of-the-art T2V frameworks with negligible computational overhead and minimal compromise in visual quality. Additionally, we demonstrate that this noise-induced contrastive mechanism is effective not only for sharpening motion trajectories but also for steering complex, non-linear concepts such as precise object numerosity, which are typically difficult to modulate via standard text-based guidance.

We address the attribution problem for apparent slow collective dynamics: is the observed persistence intrinsic, or inherited from a persistent driver? For the leading eigenvalue fraction $ψ_1=λ_{\max}/N$ of S\&P 500 60-day rolling correlation matrices ($237$ stocks, 2004--2023), a VIX-coupled Ornstein--Uhlenbeck model reduces the effective relaxation time from $298$ to $61$ trading days and improves the fit over bare mean reversion by $Δ$BIC$=109$. On the decomposition sample, an informational residual of $\log(\mathrm{VIX})$ alone retains most of that gain ($Δ$BIC$=78.6$), whereas a mechanical VIX proxy alone does not improve the fit. Autocorrelation-matched placebo fields fail ($Δ$BIC$_{\max}=2.7$), disjoint weekly reconstructions still favor the field-coupled model ($Δ$BIC$=140$--$151$), and six anchored chronological holdouts preserve the out-of-sample advantage. Quiet-regime and field-stripped residual autocorrelation controls show the same collapse of persistence. Stronger hidden-variable extensions remain only partially supported. Within the tested stochastic class, conditioning on the observed VIX proxy absorbs most of the apparent slow dynamics.

The occurrence of large-scale power outages induced by natural disasters has been on the rise in a changing climate. Such power outages often last extended durations, causing substantial financial losses and socioeconomic impacts to customers. Accurate estimation of outage duration is thus critical for enhancing the resilience of energy infrastructure under severe weather. We formulate such a task as a machine learning (ML) problem with focus on unique real-world challenges: high-order spatial dependency in the data, a moderate number of large-scale outage events, heterogeneous types of such events, and different impacts in a region within each event. To address these challenges, we develop a Bimodal Gated Graph Attention Network (BiGGAT), a graph-based neural network model, that integrates a Graph Attention Network (GAT) with a Gated Recurrent Unit (GRU) to capture the complex spatial characteristics. We evaluate the approach in a setting of inductive learning, using large-scale power outage data from six major hurricanes in the Southeastern United States. Experimental results demonstrate that BiGGAT achieves a superior performance compared to benchmark models.

Many real-world multi-party negotiations unfold as sequences of binding, action-level commitments rather than a single final outcome. We introduce a benchmark for this under-studied regime featuring a configurable game generator that sweeps key structural properties such as incentive alignment, goal complexity, and payoff distribution. To evaluate decision-making, we test three value-function approximations - myopic reward, an optimistic upper bound, and a pessimistic lower bound - that act as biased lenses on deal evaluation. Through exact evaluation on small games and comparative evaluation on large, document-grounded instances derived from the Harvard Negotiation Challenge, we map the strategic regimes where each approximation succeeds or fails. We observe that different game structures demand different valuation strategies, motivating agents that learn robust state values and plan effectively over long horizons under binding commitments and terminal only rewards.

Diffusion models are the go-to method for Text-to-Image generation, but their iterative denoising processes has high inference latency. Quantization reduces compute time by using lower bitwidths, but applies a fixed precision across all denoising timesteps, leaving an entire optimization axis unexplored. We propose TMPDiff, a temporal mixed-precision framework for diffusion models that assigns different numeric precision to different denoising timesteps. We hypothesize that quantization errors accumulate additively across timesteps, which we then validate experimentally. Based on our observations, we develop an adaptive bisectioning-based algorithm, which assigns per-step precisions with linear evaluation complexity, reducing an otherwise exponential search problem. Across four state-of-the-art diffusion models and three datasets, TMPDiff consistently outperforms uniform-precision baselines at matched speedup, achieving 10 to 20% improvement in perceptual quality. On FLUX.1-dev, TMPDiff achieves 90% SSIM relative to the full-precision model at a speedup of 2.5x over 16-bit inference.

Large language model agents demonstrate expert-level reasoning, yet consistently fail on enterprise-specific tasks due to missing domain knowledge -- terminology, operational procedures, system interdependencies, and institutional decisions that exist largely as tribal knowledge. Current approaches fall into two categories: top-down knowledge engineering, which documents domain knowledge before agents use it, and bottom-up automation, where agents learn from task experience. Both have fundamental limitations: top-down efforts produce bloated, untested knowledge bases; bottom-up approaches cannot acquire knowledge that exists only in human heads. We present Demand-Driven Context (DDC), a problem-first methodology that uses agent failure as the primary signal for what domain knowledge to curate. Inspired by Test-Driven Development, DDC inverts knowledge engineering: instead of curating knowledge and hoping it is useful, DDC gives agents real problems, lets them demand the context they need, and curates only the minimum knowledge required to succeed. We describe the methodology, its entity meta-model, and a convergence hypothesis suggesting that 20-30 problem cycles produce a knowledge base sufficient for a given domain role. We demonstrate DDC through a worked example in retail order fulfillment, where nine cycles targeting an SRE incident management agent produce a reusable knowledge base of 46 entities. Finally, we propose a scaling architecture for enterprise adoption with semi-automated curation and human governance.

Modernizing large legacy systems remains a major challenge in enterprise environments, particularly when migration must preserve domain-specific logic while conforming to internal architectural frameworks and shared APIs. Direct application of Large Language Models (LLMs) for code translation often produces syntactically valid outputs that fail to compile or integrate within existing production frameworks, limiting their practical adoption in real-world modernization efforts. In this paper, we propose LegacyTranslate, a multi-agent framework for API-aware code translation, developed and evaluated in the context of an ongoing modernization effort at a financial institution migrating approximately 2.5 million lines of PL/SQL to Java. The core idea is to use specialized LLM-based agents, each addressing a different aspect of the translation challenge. Specifically, LegacyTranslate consists of three agents: Initial Translation Agent produces an initial Java translation using retrieved in-context examples; API Grounding Agent aligns the code with existing APIs by retrieving relevant entries from an API knowledge base; and Refinement Agent iteratively refines the output using compiler feedback and API suggestions to improve correctness. Our experiments show that each agent contributes to better translation quality. The Initial Translation Agent alone achieves 45.6% compilable outputs and 30.9% test-pass rate. With API Grounding Agent and Refinement Agent, compilation improves by an additional 8% and test-pass accuracy increases by 3%.

Modern Translation Systems heavily rely on high-quality, large parallel datasets for state-of-the-art performance. However, such resources are largely unavailable for most of the South Asian languages. Among them, Nepali and Tamang fall into such category, with Tamang being among the least digitally resourced languages in the region. This work addresses the gap by developing NepTam20K, a 20K gold standard parallel corpus, and NepTam80K, an 80K synthetic Nepali-Tamang parallel corpus, both sentence-aligned and designed to support machine translation. The datasets were created through a pipeline involving data scraping from Nepali news and online sources, pre-processing, semantic filtering, balancing for tense and polarity (in NepTam20K dataset), expert translation into Tamang by native speakers of the language, and verification by an expert Tamang linguist. The dataset covers five domains: Agriculture, Health, Education and Technology, Culture, and General Communication. To evaluate the dataset, baseline machine translation experiments were carried out using various multilingual pre-trained models: mBART, M2M-100, NLLB-200, and a vanilla Transformer model. The fine-tuning on the NLLB-200 achieved the highest sacreBLEU scores of 40.92 (Nepali-Tamang) and 45.26 (Tamang-Nepali).

This paper presents a multi-agent perception-action exploration alliance, dubbed A4VL, for efficient long-video reasoning. A4VL operates in a multi-round perception-action exploration loop with a selection of VLM agents. In each round, the team of agents performs video question-answer (VideoQA) via perception exploration followed by action exploration. During perception exploration, each agent learns to extract query-specific perception clue(s) from a few sampled frames and performs clue-based alignment to find the video block(s) that are most relevant to the query-specific event. During action exploration, A4VL performs video reasoning in three steps: (1) each agent produces its initial answer with rational, (2) all agents collaboratively scores one another through cross-reviews and relevance ranking, and (3) based on whether a satisfactory consensus is reached, the decision is made either to start a new round of perception-action deliberation by pruning (e.g., filtering out the lowest performing agent) and re-staging (e.g., new-clue and matching block based perception-action exploration), or to conclude by producing its final answer. The integration of the multi-agent alliance through multi-round perception-action exploration, coupled with event-driven partitioning and cue-guided block alignment, enables A4VL to effectively scale to real world long videos while preserving high quality video reasoning. Evaluation Results on five popular VideoQA benchmarks show that A4VL outperforms 18 existing representative VLMs and 10 recent methods optimized for long-video reasoning, while achieving significantly lower inference latency. Our code is released at https://github.com/git-disl/A4VL.

We propose a society-first theory of normative appropriateness where individuals, modeled as pre-trained actors with cognitive architectures analogous to Large Language Models (LLMs), generate behavior via predictive pattern completion. Our theory posits that individuals act by completing distributed symbolic patterns based on context, answering questions such as "What does a person such as I do in a situation such as this?". This sense-making mechanism provides a parsimonious account of the key features of human norms: their context-dependence, arbitrariness, automaticity, dynamism, and their support from social sanctioning. It challenges rational-choice theories of social norms by accounting for their key features without needing to exogenously posit scalar rewards or preference relations. By distinguishing between explicit norms, which we associate with in-context adaptation, and implicit norms, which we associate with long-term memory, the theory reconceptualizes several foundational ideas in cognitive science. In particular, it gives an alternative account to the data traditionally seen as supporting dual-process models, and it flips the role of rationality, allowing us to construe it as adherence to culturally-contingent justification standards.

This work studies the Schrödinger bridge problem for the kinematic equation on a compact connected Lie group. The objective is to steer a controlled diffusion between given initial and terminal densities supported over the Lie group while minimizing the control effort. We develop a coordinate-free formulation of this stochastic optimal control problem that respects the underlying geometric structure of the Lie group, thereby avoiding limitations associated with local parameterizations or embeddings in Euclidean spaces. We establish the existence and uniqueness of solution to the corresponding Schrödinger system. Our results are constructive in that they derive a geometric controller that optimally interpolates probability densities supported over the Lie group. To illustrate the results, we provide numerical examples on $\mathsf{SO}(2)$ and $\mathsf{SO}(3)$.

Graph-RAG systems achieve strong multi-hop question answering by indexing documents into knowledge graphs, but strong retrieval does not guarantee strong answers. Evaluating KET-RAG, a leading Graph-RAG system, on three multi-hop QA benchmarks (HotpotQA, MuSiQue, 2WikiMultiHopQA), we find that 77% to 91% of questions have the gold answer in the retrieved context, yet accuracy is only 35% to 78%, and 73% to 84% of errors are reasoning failures. We propose two augmentations: (i) SPARQL chain-of-thought prompting, which decomposes questions into triple-pattern queries aligned with the entity-relationship context, and (ii) graph-walk compression, which compresses the context by ~60% via knowledge-graph traversal with no LLM calls. SPARQL CoT improves accuracy by +2 to +14 pp; graph-walk compression adds +6 pp on average when paired with structured prompting on smaller models. Surprisingly, we show that, with question-type routing, a fully augmented budget open-weight Llama-8B model matches or exceeds the unaugmented Llama-70B baseline on all three benchmarks at ~12x lower cost. A replication on LightRAG confirms that our augmentations transfer across Graph-RAG systems.

The enhancement of reasoning capabilities in large language models (LLMs) has garnered significant attention, with supervised fine-tuning (SFT) and reinforcement learning emerging as dominant paradigms. While recent studies recognize the importance of reflection in reasoning processes, existing methodologies seldom address proactive reflection encouragement during training. This study focuses on mathematical reasoning by proposing a four-stage framework integrating Group Relative Policy Optimization (GRPO) with reflection reward mechanisms to strengthen LLMs' self-reflective capabilities. Besides, this approach incorporates established accuracy and format reward. Experimental results demonstrate GRPO's state-of-the-art performance through reflection-encouraged training, with ablation studies confirming the reflection reward's pivotal role. Comparative evaluations demonstrate full-parameter SFT's superiority over low-rank adaptation (LoRA) despite heightened computational demands. Building on these cumulative findings, this research substantiates GRPO's methodological significance in post-training optimization and envisions its potential to serve as a pivotal enabler for future LLM-based intelligent agents through the synergistic integration of cognitive rewards with dynamic environmental interactions.

State-of-the-art spoken dialogue models (Défossez et al. 2024; Schalkwyk et al. 2025) use neural audio codecs to "tokenize" audio signals into a lower-frequency stream of vectorial latent representations, each quantized using a hierarchy of vector codebooks. A transformer layer allows these representations to reflect some time- and context-dependent patterns. We train probes on labeled audio data from Cole et al. (2023) to test whether the pitch trajectories that characterize English phrase-final (nuclear) intonational tunes are among these patterns. Results: Linear probes trained on the unquantized latents or some of the associated codewords yield above-chance accuracy in distinguishing eight phonologically specified nuclear tunes with monotonal pitch accents (top average test accuracy (TATA): 0.31) and the five clusters of these tunes that are robust in human speech production and perception (TATA: 0.45). Greater accuracy (TATAs: 0.74-0.89) is attained for binary distinctions between classes of rising vs. falling tunes, respectively used for questions and assertions. Information about tunes is spread among all codebooks, which calls into question a distinction between 'semantic' and 'acoustic' codebooks found in the literature. Accuracies improve with nonlinear probes, but discrimination among the five clusters remains far from human performance, suggesting a fundamental limitation of current codecs.

Audio anti-spoofing systems are typically formulated as binary classifiers distinguishing bona fide from spoofed speech. This assumption fails under layered generative processing, where benign transformations introduce distributional shifts that are misclassified as spoofing. We show that phonation-modifying voice conversion and speech restoration are treated as out-of-distribution despite preserving speaker authenticity. Using a multi-class setup separating bona fide, converted, spoofed, and converted-spoofed speech, we analyse model behaviour through self-supervised learning (SSL) embeddings and acoustic correlates. The benign transformations induce a drift in the SSL space, compressing bona fide and spoofed speech and reducing classifier separability. Reformulating anti-spoofing as a multi-class problem improves robustness to benign shifts while preserving spoof detection, suggesting binary systems model the distribution of raw speech rather than authenticity itself.

A task-specific model trained on 212,231 UK Biobank subjects to predict vascular age from PPG (AI-PPG Age) fails on a different clinical population: predictions collapse to a narrow 38-67 year range regardless of true age. Meanwhile, a general-purpose foundation model with no age-related training objective achieves lower error on the same data. We investigate why this happens and what it means for PPG-based biological age prediction. We evaluate three open-source PPG models (Pulse-PPG, PaPaGei-S, AI-PPG Age) on 906 surgical patients from PulseDB, using frozen embeddings with Ridge regression and 5-fold cross-validation. Pulse-PPG reaches MAE = 9.28 years, beating both AI-PPG Age in linear probe mode (9.72) and HR/HRV combined with demographics (9.59). Adding demographic features brings the best result down to MAE = 8.22 years (R2 = 0.517, r = 0.725). The predicted age gap correlates with diastolic blood pressure after adjusting for chronological age (r = -0.188, p = 1.2e-8), consistent with what Apple reported for their proprietary PpgAge model. The remaining gap with Apple (MAE 2.43) appears driven by dataset size (906 vs 213,593 subjects) and population differences rather than model architecture, as our learning curve shows no plateau. Code is publicly available.

A hybrid digital twin framework is presented for bridge condition monitoring using existing traffic cameras and weather APIs, reducing reliance on dedicated sensor installations. The approach is demonstrated on the Peace Bridge (99 years in service) under high traffic demand and harsh winter exposure. The framework fuses three near-real-time streams: YOLOv8 computer vision from a bridge-deck camera estimates vehicle counts, traffic density, and load proxies; a Lighthill--Whitham--Richards (LWR) model propagates density $ρ(x,t)$ and detects deceleration-driven shockwaves linked to repetitive loading and fatigue accumulation; and weather APIs provide deterioration drivers including temperature cycling, freeze-thaw activity, precipitation-related corrosion potential, and wind effects. Monte Carlo simulation quantifies uncertainty across traffic-environment scenarios, while Random Forest models map fused features to fatigue indicators and maintenance classification. The framework demonstrates utilizing existing infrastructure for cost-effective predictive maintenance of aging, high-traffic bridges in harsh climates.

Political speakers often avoid answering questions directly while maintaining the appearance of responsiveness. Despite its importance for public discourse, such strategic evasion remains underexplored in Natural Language Processing. We introduce SemEval-2026 Task 6, CLARITY, a shared task on political question evasion consisting of two subtasks: (i) clarity-level classification into Clear Reply, Ambivalent, and Clear Non-Reply, and (ii) evasion-level classification into nine fine-grained evasion strategies. The benchmark is constructed from U.S. presidential interviews and follows an expert-grounded taxonomy of response clarity and evasion. The task attracted 124 registered teams, who submitted 946 valid runs for clarity-level classification and 539 for evasion-level classification. Results show a substantial gap in difficulty between the two subtasks: the best system achieved 0.89 macro-F1 on clarity classification, surpassing the strongest baseline by a large margin, while the top evasion-level system reached 0.68 macro-F1, matching the best baseline. Overall, large language model prompting and hierarchical exploitation of the taxonomy emerged as the most effective strategies, with top systems consistently outperforming those that treated the two subtasks independently. CLARITY establishes political response evasion as a challenging benchmark for computational discourse analysis and highlights the difficulty of modeling strategic ambiguity in political language.

Part-level 3D generation is crucial for various downstream applications, including gaming, film production, and industrial design. However, decomposing a 3D shape into geometrically plausible and meaningful components remains a significant challenge. Previous part-based generation methods often struggle to produce well-constructed parts, exhibiting poor structural coherence, geometric implausibility, inaccuracy, or inefficiency. To address these challenges, we introduce EI-Part, a novel framework specifically designed to generate high-quality 3D shapes with components, characterized by strong structural coherence, geometric plausibility, geometric fidelity, and generation efficiency. We propose utilizing distinct representations at different stages: an Explode state for part completion and an Implode state for geometry refinement. This strategy fully leverages spatial resolution, enabling flexible part completion and fine geometric detail generation. To maintain structural coherence between parts, a self-attention mechanism is incorporated in both exploded and imploded states, facilitating effective information perception and feature fusion among components during generation. Extensive experiments on multiple benchmarks demonstrate that EI-Part efficiently produces semantically meaningful and structurally coherent parts with fine-grained geometric details, achieving state-of-the-art performance in part-level 3D generation. Project page: https://cvhadessun.github.io/EI-Part/

Hash-based maps, particularly java.util.HashMap, are pervasive in Java applications and the JVM, making their performance critical. Evaluating optimizations is challenging because performance depends on factors such as operation patterns, key distributions, and resizing behavior. Microbenchmarks are fast and repeatable but often oversimplify workloads, failing to capture the realistic usage patterns. Application benchmarks (e.g., DaCapo, Renaissance) provide realistic usages but are more expensive to run, prone to variability, and dominated by non-HashMap computations, making map-related performance changes difficult to observe. To address this challenge, we propose MapReplay, a benchmarking methodology that combines the realism of application benchmarks with the efficiency of microbenchmarks. MapReplay traces HashMap API usages generating a replay workload that reproduces the same operation sequence while faithfully reconstructing internal map states. This enables realistic and efficient evaluation of alternative implementations under realistic usage patterns. Applying MapReplay to DaCapo-Chopin and Renaissance, the resulting suite, MapReplayBench, reproduces application-level performance trends while reducing experimentation time and revealing insights difficult to obtain from full benchmarks.

We introduce Aumann-SHAP, an interaction-aware framework that decomposes counterfactual transitions by restricting the model to a local hypercube connecting baseline and counterfactual features. Each hyper-cube is decomposed into a grid in order to construct an induced micro-player cooperative game in which elementary grid-step moves become players. Shapley and LES values on this TU-micro-game yield: (i) within-pot contribution of each feature to the interaction with other features (interaction explainability), and (ii) the contribution of each instance and each feature to the counterfactual analysis (individual and global explainability). In particular, Aumann-LES values produce individual and global explanations along the counterfactual transition. Shapley and LES values converge to the diagonal Aumann-Shapley (integrated-gradients) attribution method. Experiments on the German Credit dataset and MNIST data show that Aumann-LES produces robust results and better explanations than the standard Shapley value during the counterfactual transition.

This work proposes a formal abductive explanation framework designed to systematically uncover rationales underlying AI predictions of mental health help-seeking within tech workplace settings. By computing rigorous justifications for model outputs, this approach enables principled selection of models tailored to distinct psychiatric profiles and underpins ethically robust recourse planning. Beyond moving past ad-hoc interpretability, we explicitly examine the influence of sensitive attributes such as gender on model decisions, a critical component for fairness assessments. In doing so, it aligns explanatory insights with the complex landscape of workplace mental health, ultimately supporting trustworthy deployment and targeted interventions.

GraphRAG is increasingly adopted for converting unstructured corpora into graph structures to enable multi-hop reasoning. However, standard graph algorithms rely heavily on static connectivity and explicit edges, often failing in real-world scenarios where knowledge graphs (KGs) are noisy, sparse, or incomplete. To address this limitation, we introduce INSES (Intelligent Navigation and Similarity Enhanced Search), a dynamic framework designed to reason beyond explicit edges. INSES couples LLM-guided navigation, which prunes noise and steers exploration, with embedding-based similarity expansion to recover hidden links and bridge semantic gaps. Recognizing the computational cost of graph reasoning, we complement INSES with a lightweight router that delegates simple queries to Naïve RAG and escalates complex cases to INSES, balancing efficiency with reasoning depth. INSES consistently outperforms SOTA RAG and GraphRAG baselines across multiple benchmarks. Notably, on the MINE benchmark, it demonstrates superior robustness across KGs constructed by varying methods (KGGEN, GraphRAG, OpenIE), improving accuracy by 5%, 10%, and 27%, respectively.

Latent space-based facial attribute editing methods have gained popularity in applications such as digital entertainment, virtual avatar creation, and human-computer interaction systems due to their potential for efficient and flexible attribute manipulation, particularly for continuous edits. Among these, unsupervised latent space-based methods, which discover effective semantic vectors without relying on labeled data, have attracted considerable attention in the research community. However, existing methods still encounter difficulties in disentanglement, as manipulating a specific facial attribute may unintentionally affect other attributes, complicating fine-grained controllability. To address these challenges, we propose a novel framework designed to offer an effective and adaptable solution for unsupervised facial attribute editing, called Unsupervised Facial Attribute Controllable Editing (U-Face). The proposed method frames semantic vector learning as a subspace learning problem, where latent vectors are approximated within a lower-dimensional semantic subspace spanned by a semantic vector matrix. This formulation can also be equivalently interpreted from a projection-reconstruction perspective and further generalized into an autoencoder framework, providing a foundation that can support disentangled representation learning in a flexible manner. To improve disentanglement and controllability, we impose orthogonal non-negative constraints on the semantic vectors and incorporate attribute boundary vectors to reduce entanglement in the learned directions. Although these constraints make the optimization problem challenging, we design an alternating iterative algorithm, called Alternating Iterative Disentanglement and Controllability (AIDC), with closed-form updates and provable convergence under specific conditions.

The widespread adoption of electronic health records (EHRs) enables the acquisition of heterogeneous clinical data, spanning lab tests, vital signs, medications, and procedures, which offer transformative potential for artificial intelligence in healthcare. Although traditional modelling approaches have typically relied on multivariate time series, they often struggle to accommodate the inherent sparsity and irregularity of real-world clinical workflows. Consequently, research has shifted toward event stream representation, which treats patient records as continuous sequences, thereby preserving the precise temporal structure of the patient journey. However, the existing literature remains fragmented, characterised by inconsistent definitions, disparate modelling architectures, and varying training protocols. To address these gaps, this review establishes a unified definition of EHR event streams and introduces a novel taxonomy that categorises models based on their handling of event time, type, and value. We systematically review training strategies, ranging from supervised learning to self-supervised methods, and provide a comprehensive discussion of applications across clinical scenarios. Finally, we identify open critical challenges and future directions, with the aim of clarifying the current landscape and guiding the development of next-generation healthcare models.

Requirements traceability in safety-critical software development remains largely dependent on external documentation maintained separately from the systems it describes. This separation introduces structural fragility: traces degrade silently as requirements, code, and tests evolve independently across tools, repositories, and revisions. Recent advances in LLM-based traceability focus on recovering broken links after the fact - an inherently retrospective approach. This paper introduces ReqToCode, an approach that prevents trace degradation by embedding traceable system elements directly into the codebase, making traceability a compile-time verifiable property of the system rather than an external documentation task. Central to the approach is the concept of the Traceable - a language-native, generated code element that represents a single requirement and carries its metadata. Developers reference Traceables in implementation and test code, creating hard, bidirectional links that are validated automatically during the build process. When requirements change, the system responds through a graduated lifecycle - from deprecation warnings to build failures - providing teams with actionable signals rather than abrupt breakage. We describe the approach, its architectural principles, the Traceable lifecycle, and illustrate it with a generic example spanning requirement definition, artifact generation, code integration, and build-time validation.

While graph-derived signals are widely used in tabular learning, existing studies typically rely on limited experimental setups and average performance comparisons, leaving the statistical reliability and robustness of observed gains largely unexplored. Consequently, it remains unclear which signals provide consistent and robust improvements. This paper presents a taxonomy-driven empirical analysis of graph-derived signals for tabular machine learning. We propose a unified and reproducible evaluation protocol to systematically assess which categories of graph-derived signals yield statistically significant and robust performance improvements. The protocol provides an extensible setup for the controlled integration of diverse graph-derived signals into tabular learning pipelines. To ensure a fair and rigorous comparison, it incorporates automated hyperparameter optimization, multi-seed statistical evaluation, formal significance testing, and robustness analysis under graph perturbations. We demonstrate the protocol through an extensive case study on a large-scale, imbalanced cryptocurrency fraud detection dataset. The analysis identifies signal categories providing consistently reliable performance gains and offers interpretable insights into which graph-derived signals indicate fraud-discriminative structural patterns. Furthermore, robustness analyses reveal pronounced differences in how various signals handle missing or corrupted relational data. These findings demonstrate practical utility for fraud detection and illustrate how the proposed taxonomy-driven evaluation protocol can be applied in other application domains.

Web search has become an inevitable part of everyday life. Improving and monetizing web search has been a focus of major Internet players. Understanding the context of web search query is an important aspect of this task as it represents unobserved facts that add meaning to an otherwise incomplete query.The context of a query consists of user's location, local time, search history, behavioral segments, installed apps on their phone and so on. Queries that either explicitly use location context (eg: "best hotels in New York City") or implicitly refer to the user's physical location (e.g. "coffee shops near me") are becoming increasingly common on mobile devices. Understanding and representing the user's interest location and/or physical location is essential for providing a relevant user experience. In this study, we developed a simple and powerful neural embedding based framework to represent a user's query and their location in a single low-dimensional space. We show that this representation is able to capture the subtle interactions between the user's query intent and query/physical location, while improving the ad ranking and query-ad relevance scores over other location-unaware approaches and location-aware approaches.

Vision foundation models trained with self-supervised objectives achieve strong performance across diverse tasks and exhibit emergent object segmentation properties. However, their alignment with human object perception remains poorly understood. Here, we introduce a behavioral benchmark in which participants make same/different object judgments for dot pairs on naturalistic scenes, scaling up a classical psychophysics paradigm to over 1000 trials. We test a diverse set of vision models using a simple readout from their representations to predict subjects' reaction times. We observe a steady improvement across model generations, with both architecture and training objective contributing to alignment, and transformer-based models trained with the DINO self-supervised objective showing the strongest performance. To investigate the source of this improvement, we propose a novel metric to quantify the object-centric component of representations by measuring patch similarity within and between objects. Across models, stronger object-centric structure predicts human segmentation behavior more accurately. We further show that matching the Gram matrix of supervised transformer models, capturing similarity structure across image patches, with that of a self-supervised model through distillation improves their alignment with human behavior, converging with the prior finding that Gram anchoring improves DINOv3's feature quality. Together, these results demonstrate that self-supervised vision models capture object structure in a behaviorally human-like manner, and that Gram matrix structure plays a role in driving perceptual alignment.

Space missions generate massive volumes of high-resolution orbital and surface imagery that far exceed the capacity for manual inspection. Detecting rare phenomena is scientifically critical, yet traditional supervised learning struggles due to scarce labeled examples and closed-world assumptions that prevent discovery of genuinely novel observations. In this work, we investigate Visual Anomaly Detection (VAD) as a framework for automated discovery in planetary exploration. We present the first empirical evaluation of state-of-the-art feature-based VAD methods on real planetary imagery, encompassing both orbital lunar data and Mars rover surface imagery. To support this evaluation, we introduce two benchmarks: (i) a lunar dataset derived from Lunar Reconnaissance Orbiter Camera Narrow Angle imagery, comprising of fresh and degraded craters as anomalies alongside normal terrain; and (ii) a Mars surface dataset designed to reflect the characteristics of rover-acquired imagery. We evaluate multiple VAD approaches with a focus on computationally efficient, edge-oriented solutions suitable for onboard deployment, applicable to both orbital platforms surveying the lunar surface and surface rovers operating on Mars. Our results demonstrate that feature-based VAD methods can effectively identify rare planetary surface phenomena while remaining feasible for resource-constrained environments. By grounding anomaly detection in planetary science, this work establishes practical benchmarks and highlights the potential of open-world perception systems to support a range of mission-critical applications, including tactical planning, landing site selection, hazard detection, bandwidth-aware data prioritization, and the discovery of unanticipated geological processes.

Closed-source large language models (LLMs), such as ChatGPT and Gemini, are increasingly consulted for medical advice, yet their explanations may appear plausible while failing to reflect the model's underlying reasoning process. This gap poses serious risks as patients and clinicians may trust coherent but misleading explanations. We conduct a systematic black-box evaluation of faithfulness in medical reasoning among three widely used closed-source LLMs. Our study consists of three perturbation-based probes: (1) causal ablation, testing whether stated chain-of-thought (CoT) reasoning causally influences predictions; (2) positional bias, examining whether models create post-hoc justifications for answers driven by input positioning; and (3) hint injection, testing susceptibility to external suggestions. We complement these quantitative probes with a small-scale human evaluation of model responses to patient-style medical queries to examine concordance between physician assessments of explanation faithfulness and layperson perceptions of trustworthiness. We find that CoT reasoning steps often do not causally drive predictions, and models readily incorporate external hints without acknowledgment. In contrast, positional biases showed minimal impact in this setting. These results underscore that faithfulness, not just accuracy, must be central in evaluating LLMs for medicine, to ensure both public protection and safe clinical deployment.

Pre-trained Large Language Model (LLM) exhibits broad capabilities, yet, for specific tasks or domains their attainment of higher accuracy and more reliable reasoning generally depends on post-training through Supervised Fine-Tuning (SFT) or Reinforcement Learning (RL). Although often treated as distinct methodologies, recent theoretical and empirical developments demonstrate that SFT and RL are closely connected. This study presents a comprehensive and unified perspective on LLM post-training with SFT and RL. We first provide an in-depth overview of both techniques, examining their objectives, algorithmic structures, and data requirements. We then systematically analyze their interplay, highlighting frameworks that integrate SFT and RL, hybrid training pipelines, and methods that leverage their complementary strengths. Drawing on a representative set of recent application studies from 2023 to 2025, we identify emerging trends, characterize the rapid shift toward hybrid post-training paradigms, and distill key takeaways that clarify when and why each method is most effective. By synthesizing theoretical insights, practical methodologies, and empirical evidence, this study establishes a coherent understanding of SFT and RL within a unified framework and outlines promising directions for future research in scalable, efficient, and generalizable LLM post-training.

Recurrent Neural Networks (RNNs) are vital for sequential data processing. Long Short-Term Memory Autoencoders (LSTM-AEs) are particularly effective for unsupervised anomaly detection in time-series data. However, inherent sequential dependencies limit parallel computation. While previous work has explored FPGA-based acceleration for LSTM networks, efforts have typically focused on optimizing a single LSTM layer at a time. We introduce a novel FPGA-based accelerator using a dataflow architecture that exploits temporal parallelism for concurrent multi-layer processing of different timesteps within sequences. Experimental evaluations on four representative LSTM-AE models with varying widths and depths, implemented on a Zynq UltraScale+ MPSoC FPGA, demonstrate significant advantages over CPU (Intel Xeon Gold 5218R) and GPU (NVIDIA V100) implementations. Our accelerator achieves latency speedups up to 79.6x vs. CPU and 18.2x vs. GPU, alongside energy-per-timestep reductions of up to 1722x vs. CPU and 59.3x vs. GPU. These results, including superior network depth scalability, highlight our approach's potential for high-performance, real-time, power-efficient LSTM-AE-based anomaly detection on FPGAs.

Modern large language model training is no longer limited by data availability, but by the inability of existing preprocessing pipelines to simultaneously achieve massive scale and high data quality. Current approaches are forced to sacrifice one for the other: either aggressively filtering to improve quality at the cost of severe token loss, or retaining large volumes of data while introducing substantial noise. In this work, we introduce FLUX, a preprocessing pipeline specifically designed to break this long-standing trade-off by maximizing token retention while enforcing rigorous quality control. Models trained on FLUX-curated data consistently outperform prior methods. A 3B-parameter model trained on 60B tokens with FLUX achieves 32.14% MMLU accuracy, surpassing the previous state-of-the-art pipeline DCLM (31.98%) and significantly outperforming FineWeb (29.88%). FLUX achieves the same aggregate score as a model trained on DCLM data using only 39B tokens, resulting in a 34.4% reduction in training compute. At the data level, FLUX extracts 50B usable tokens from a single dump (CC-MAIN-2025-51), compared to 40B from DCLM (+25% retention). FLUX-Base yields 192B tokens, exceeding FineWeb's 170B while still maintaining superior quality. Overall, FLUX establishes a new state of the art in web-scale data preprocessing by demonstrating that high retention, strong quality control, and computational efficiency can be achieved simultaneously, redefining the limits of scalable dataset construction for modern language models.

In offline reinforcement learning (RL), single-step temporal-difference (TD) learning can suffer from bootstrapping error accumulation over long horizons. Action-chunked TD methods mitigate this by backing up over multiple steps, but can introduce suboptimality by restricting the policy class to open-loop action sequences. To resolve this trade-off, we present Chunk-Guided Q-Learning (CGQ), a single-step TD algorithm that guides a fine-grained single-step critic by regularizing it toward a chunk-based critic trained using temporally extended backups. This reduces compounding error while preserving fine-grained value propagation. We theoretically show that CGQ attains tighter critic optimality bounds than either single-step or action-chunked TD learning alone. Empirically, CGQ achieves strong performance on challenging long-horizon OGBench tasks, often outperforming both single-step and action-chunked methods.

In High Energy Physics, as in many other fields of science, the application of machine learning techniques has been crucial in advancing our understanding of fundamental phenomena. Increasingly, deep learning models are applied to analyze both simulated and experimental data. In most experiments, a rigorous regime of testing for physically motivated systematic uncertainties is in place. The numerical evaluation of these tests for differences between the data on the one side and simulations on the other side quantifies the effect of potential sources of mismodelling on the machine learning output. In addition, thorough comparisons of marginal distributions and (linear) feature correlations between data and simulation in "control regions" are applied. However, the guidance by physical motivation, and the need to constrain comparisons to specific regions, does not guarantee that all possible sources of deviations have been accounted for. We therefore propose a new adversarial attack - the CONSERVAttack - designed to exploit the remaining space of hypothetical deviations between simulation and data after the above mentioned tests. The resulting adversarial perturbations are consistent within the uncertainty bounds - evading standard validation checks - while successfully fooling the underlying model. We further propose strategies to mitigate such vulnerabilities and argue that robustness to adversarial effects must be considered when interpreting results from deep learning in particle physics.

Echocardiography is widely used for assessing cardiac function, where clinically meaningful parameters such as left-ventricular ejection fraction (EF) play a central role in diagnosis and management. Generative models capable of synthesising realistic echocardiogram videos with explicit control over such parameters are valuable for data augmentation, counterfactual analysis, and specialist training. However, existing approaches typically rely on computationally expensive multi-step sampling and aggressive temporal normalisation, limiting efficiency and applicability to heterogeneous real-world data. We introduce EchoLVFM, a one-step latent video flow-matching framework for controllable echocardiogram generation. Operating in the latent space, EchoLVFM synthesises temporally coherent videos in a single inference step, achieving a $\mathbf{\sim 50\times}$ improvement in sampling efficiency compared to multi-step flow baselines while maintaining visual fidelity. The model supports global conditioning on clinical variables, demonstrated through precise control of EF, and enables reconstruction and counterfactual generation from partially observed sequences. A masked conditioning strategy further removes fixed-length constraints, allowing shorter sequences to be retained rather than discarded. We evaluate EchoLVFM on the CAMUS dataset under challenging single-frame conditioning. Quantitative and qualitative results demonstrate competitive video quality, strong EF adherence, and 57.9% discrimination accuracy by expert clinicians which is close to chance. These findings indicate that efficient, one-step flow matching can enable practical, controllable echocardiogram video synthesis without sacrificing fidelity. Code available at: https://github.com/EngEmmanuel/EchoLVFM

Vision Language Action VLA models are typically evaluated using per benchmark scripts maintained independently by each model repository, leading to duplicated code, dependency conflicts, and underspecified protocols. We present vla eval, an open source evaluation harness that decouples model inference from benchmark execution through a WebSocket msgpack protocol with Docker based environment isolation. Models integrate once by implementing a single predict() method; benchmarks integrate once via a four method interface; the full cross evaluation matrix works automatically. A complete evaluation requires only two commands: vla eval serve and vla eval run. The framework supports 13 simulation benchmarks and six model servers. Parallel evaluation via episode sharding and batch inference achieves a 47x throughput improvement, completing 2000 LIBERO episodes in about 18 minutes. Using this infrastructure, we conduct a reproducibility audit of a published VLA model across three benchmarks, finding that all three closely reproduce published values while uncovering undocumented requirements ambiguous termination semantics and hidden normalization statistics that can silently distort results. We additionally release a VLA leaderboard aggregating 657 published results across 17 benchmarks. Framework, evaluation configs, and all reproduction results are publicly available.

Combinatorial Testing (CT) is essential for detecting interaction-triggered faults, yet generating minimal Covering Arrays under complex constraints remains an unresolved NP-hard challenge. Current greedy algorithms are highly scalable but suffer from severe ``diminishing returns'': they efficiently cover initial interactions but produce bloated, redundant test suites when struggling to pack the final few difficult pairs. While exact mathematical programming could theoretically address this inefficiency, it has historically been intractable due to combinatorial explosion. In this paper, we pioneer the application of exact mathematical modeling to CT by introducing SeqTG, a scalable framework based on Sequential Integer Linear Programming (ILP). To circumvent the scalability barrier, SeqTG employs a novel Warm-Start strategy: a rapid greedy initialization first clears the ``easy'' interactions, allowing the rigorous ILP solver to exclusively optimize the fragmented, difficult-to-cover remainder. The pipeline operates in three stages: (1) a Constraint-First phase grouping must-include requirements via graph partitioning; (2) an Incremental Optimization phase targeting the remaining interactions with sequential ILP; and (3) a Global Minimization phase eliminating redundancies via set-covering. Extensive evaluations across standard benchmarks and 200 large-scale configurations validate the framework's efficacy. The results demonstrate that SeqTG effectively eradicates late-stage bloat, achieving state-of-the-art test suite compactness and strict constraint adherence.

Clinical question answering over electronic health records (EHRs) can help clinicians and patients access relevant medical information more efficiently. However, many recent approaches rely on large cloud-based models, which are difficult to deploy in clinical environments due to privacy constraints and computational requirements. In this work, we investigate how far grounded EHR question answering can be pushed when restricted to a single notebook. We participate in all four subtasks of the ArchEHR-QA 2026 shared task and evaluate several approaches designed to run on commodity hardware. All experiments are conducted locally without external APIs or cloud infrastructure. Our results show that such systems can achieve competitive performance on the shared task leaderboards. In particular, our submissions perform above average in two subtasks, and we observe that smaller models can approach the performance of much larger systems when properly configured. These findings suggest that privacy-preserving EHR QA systems running fully locally are feasible with current models and commodity hardware. The source code is available at https://github.com/ibrahimey/ArchEHR-QA-2026.

Automated radiology report generation holds immense potential to alleviate the heavy workload of radiologists. Despite the formidable vision-language capabilities of recent Multimodal Large Language Models (MLLMs), their clinical deployment is severely constrained by inherent limitations: their "black-box" decision-making renders the generated reports untraceable due to the lack of explicit visual evidence to support the diagnosis, and they struggle to access external domain knowledge. To address these challenges, we propose the Evidence-driven Radiology Report Generation Agent (EviAgent). Unlike opaque end-to-end paradigms, EviAgent coordinates a transparent reasoning trajectory by breaking down the complex generation process into granular operational units. We integrate multi-dimensional visual experts and retrieval mechanisms as external support modules, endowing the system with explicit visual evidence and high-quality clinical priors. Extensive experiments on MIMIC-CXR, CheXpert Plus, and IU-Xray datasets demonstrate that EviAgent outperforms both large-scale generalist models and specialized medical models, providing a robust and trustworthy solution for automated radiology report generation.

In existing Audio-Visual Speech Enhancement (AVSE) methods, objectives such as Scale-Invariant Signal-to-Noise Ratio (SI-SNR) and Mean Squared Error (MSE) are widely used; however, they often correlate poorly with perceptual quality and provide limited interpretability for optimization. This work proposes a reinforcement learning-based AVSE framework with a Large Language Model (LLM)-based interpretable reward model. An audio LLM generates natural language descriptions of enhanced speech, which are converted by a sentiment analysis model into a 1-5 rating score serving as the PPO reward for fine-tuning a pretrained AVSE model. Compared with scalar metrics, LLM-generated feedback is semantically rich and explicitly describes improvements in speech quality. Experiments on the 4th COG-MHEAR AVSE Challenge (AVSEC-4) dataset show that the proposed method outperforms a supervised baseline and a DNSMOS-based RL baseline in PESQ, STOI, neural quality metrics, and subjective listening tests.

Large Language Model (LLM) agents increasingly use external tools for complex tasks and rely on embedding-based retrieval to select a small top-k subset for reasoning. As these systems scale, the robustness of this retrieval stage is underexplored, even though prior work has examined attacks on tool selection. This paper introduces ToolFlood, a retrieval-layer attack on tool-augmented LLM agents. Rather than altering which tool is chosen after retrieval, ToolFlood overwhelms retrieval itself by injecting a few attacker-controlled tools whose metadata is carefully placed by exploiting the geometry of embedding space. These tools semantically span many user queries, dominate the top-k results, and push all benign tools out of the agent's context. ToolFlood uses a two-phase adversarial tool generation strategy. It first samples subsets of target queries and uses an LLM to iteratively generate diverse tool names and descriptions. It then runs an iterative greedy selection that chooses tools maximizing coverage of remaining queries in embedding space under a cosine-distance threshold, stopping when all queries are covered or a budget is reached. We provide theoretical analysis of retrieval saturation and show on standard benchmarks that ToolFlood achieves up to a 95% attack success rate with a low injection rate (1% in ToolBench). The code will be made publicly available at the following link: https://github.com/as1-prog/ToolFlood

Near term quantum processors operate in a noise dominated regime, motivating error mitigation techniques that recover accurate expectation values without full fault tolerance. Zero Noise Extrapolation (ZNE) is a widely used but biased error mitigation method that lacks rigorous error bounds. Its effective application requires nontrivial technical choices, most notably the selection of noise scaling factors and extrapolation models, making ZNE sensitive to user expertise and often necessitating costly trial and error procedures. Here, we introduce Folding Free Zero Noise Extrapolation (FF-ZNE), a method that removes the need for noise factor selection by achieving effective noise amplification without circuit folding. FF-ZNE exploits isomorphic hardware layouts with distinct native noise profiles, such that executing a fixed circuit across these layouts induces controllable variations in the effective noise strength. Under a depolarizing noise model, we analytically show that the resulting extrapolation admits a fixed linear form, eliminating extrapolator choice and enabling a seamless, user independent mitigation procedure. We further propose two algorithms that identify sets of isomorphic hardware layouts on which a given circuit yields sufficiently distinct expectation values to enable reliable zero noise extrapolation. Experiments on a 133 qubit IBM Quantum device demonstrate that FF-ZNE yields mitigated expectation values with average deviations of ~6% and 4.5% for up to 50 qubit EfficientSU2 (sparse) and Hamiltonian simulation (dense) circuits, respectively. The method is thus scalable and applicable to a broad range of circuits. By eliminating noise factor and extrapolator selection, FF-ZNE transforms zero noise extrapolation from a technique requiring expert tuning into a practical, scalable, and broadly accessible error mitigation method for current quantum hardware.

Standard transport protocols like TCP operate as a blind, FIFO conveyor belt for data, a model that is increasingly suboptimal for latency-sensitive and interactive applications. This paper challenges this model by introducing CATS (Conductor-driven Asymmetric Transport Scheme), a framework that provides TCP with the semantic awareness necessary to prioritize critical content. By centralizing scheduling intelligence in a transport-native "Conductor", CATS significantly improves user-perceived performance by delivering essential data first. This architecture directly confronts a cascade of historical performance workarounds and their limitations, including the high overhead of parallel connections in HTTP/1.1, the transport-layer Head-of-Line blocking in HTTP/2, and the observed implementation heterogeneity of prioritization in HTTP/3 over QUIC. Built upon TCP BBR, our ns-3 implementation demonstrates this principle by reducing the First Contentful Paint by over 78% in a representative webpage download configured as a deliberate worst-case scenario, with no penalty to total page load time compared to the baseline.

Predicting satellite imagery requires a balance between structural accuracy and textural detail. Standard deterministic methods like PredRNN or SimVP minimize pixel-based errors but suffer from the "regression to the mean" problem, producing blurry outputs that obscure subtle geographic-spatial features. Generative models provide realistic textures but often misleadingly reveal structural anomalies. To bridge this gap, we introduce Sat-JEPA-Diff, which combines Self-Supervised Learning (SSL) with Hidden Diffusion Models (LDM). An IJEPA module predicts stable semantic representations, which then route a frozen Stable Diffusion backbone via a lightweight cross-attention adapter. This ensures that the synthesized high-accuracy textures are based on absolutely accurate structural predictions. Evaluated on a global Sentinel-2 dataset, Sat-JEPA-Diff excels at resolving sharp boundaries. It achieves leading perceptual scores (GSSIM: 0.8984, FID: 0.1475) and significantly outperforms deterministic baselines, despite standard autoregressive stability limits. The code and dataset are publicly available on https://github.com/VU-AIML/SAT-JEPA-DIFF.

While large language model-based agents demonstrate great potential in collaborative tasks, their interactivity also introduces security vulnerabilities. In this paper, we propose and model group collusive attacks, a highly destructive threat in which multiple agents coordinate via sociological strategies to mislead the system. To address this challenge, we introduce GroupGuard, a training-free defense framework that employs a multi-layered defense strategy, including continuous graph-based monitoring, active honeypot inducement, and structural pruning, to identify and isolate collusive agents. Experimental results across five datasets and four topologies demonstrate that group collusive attacks increase the attack success rate by up to 15\% compared to individual attacks. GroupGuard consistently achieves high detection accuracy (up to 88\%) and effectively restores collaborative performance, providing a robust solution for securing multi-agent systems.

Recommendation systems aim to learn user interests from historical behaviors and deliver relevant items. Recent methods leverage large language models (LLMs) to construct and integrate semantic representations of users and items for capturing user interests. However, user behavior theories suggest that truly understanding user interests requires not only semantic integration but also semantic reasoning from explicit individual interests to implicit group interests. To this end, we propose an Iterative Semantic Reasoning Framework (ISRF) for generative recommendation. ISRF leverages LLMs to bridge explicit individual interests and implicit group interests in three steps. First, we perform multi-step bidirectional reasoning over item attributes to infer semantic item features and build a semantic interaction graph capturing users' explicit interests. Second, we generate semantic user features based on the semantic item features and construct a similarity-based user graph to infer the implicit interests of similar user groups. Third, we adopt an iterative batch optimization strategy, where individual explicit interests directly guide the refinement of group implicit interests, while group implicit interests indirectly enhance individual modeling. This iterative process ensures consistent and progressive interest reasoning, enabling more accurate and comprehensive user interest learning. Extensive experiments on the Sports, Beauty, and Toys datasets demonstrate that ISRF outperforms state-of-the-art baselines. The code is available at https://github.com/htired/ISRF.

Ensuring the safety and compliance of large language models (LLMs) is of paramount importance. However, existing LLM safety datasets often rely on ad-hoc taxonomies for data generation and suffer from a significant shortage of rule-grounded, real-world cases that are essential for robustly protecting LLMs. In this work, we address this critical gap by constructing a comprehensive safety dataset from a compliance perspective. Using a powerful web-searching agent, we collect a rule-grounded, real-world case dataset OmniCompliance-100K, sourced from multi-domain authoritative references. The dataset spans 74 regulations and policies across a wide range of domains, including security and privacy regulations, content safety and user data privacy policies from leading AI companies and social media platforms, financial security requirements, medical device risk management standards, educational integrity guidelines, and protections of fundamental human rights. In total, our dataset contains 12,985 distinct rules and 106,009 associated real-world compliance cases. Our analysis confirms a strong alignment between the rules and their corresponding cases. We further conduct extensive benchmarking experiments to evaluate the safety and compliance capabilities of advanced LLMs across different model scales. Our experiments reveal several interesting findings that have great potential to offer valuable insights for future LLM safety research.

Low-precision neural network training has emerged as a promising direction for reducing computational costs and democratizing access to deep learning research. However, existing 4-bit quantization methods either rely on expensive GPU infrastructure or suffer from significant accuracy degradation. In this work, we present a practical method for training convolutional neural networks at true 4-bit precision using standard PyTorch operations on commodity CPUs. We introduce a novel tanh-based soft weight clipping technique that, combined with symmetric quantization, dynamic per-layer scaling, and straight-through estimators, achieves stable convergence and competitive accuracy. Training a VGG-style architecture with 3.25 million parameters from scratch on CIFAR-10, our method achieves 92.34% test accuracy on Google Colab's free CPU tier -- matching full-precision baseline performance (92.5%) with only a 0.16% gap. We further validate on CIFAR-100, achieving 70.94% test accuracy across 100 classes with the same architecture and training procedure, demonstrating that 4-bit training from scratch generalizes to harder classification tasks. Both experiments achieve 8x memory compression over FP32 while maintaining exactly 15 unique weight values per layer throughout training. We additionally validate hardware independence by demonstrating rapid convergence on a consumer mobile device (OnePlus 9R), achieving 83.16% accuracy in only 6 epochs. To the best of our knowledge, no prior work has demonstrated 4-bit quantization-aware training achieving full-precision parity on standard CPU hardware without specialized kernels or post-training quantization.

Traditional discriminative computer vision relies predominantly on static projections, mapping input features to outputs in a single computational step. Although efficient, this paradigm lacks the iterative refinement and robustness inherent in biological vision and modern generative modelling. In this paper, we propose Discriminative Flow Matching, a framework that reformulates classification and object detection as a conditional transport process. By learning a vector field that continuously transports samples from a simple noise distribution toward a task-aligned target manifold -- such as class embeddings or bounding box coordinates -- we are at the interface between generative and discriminative learning. Our method attaches multiple independent flow predictors to a shared backbone. These predictors are trained using local flow matching objectives, where gradients are computed independently for each block. We formulate this approach for standard image classification and extend it to the complex task of object detection, where targets are high-dimensional and spatially distributed. This architecture provides the flexibility to update blocks either sequentially to minimise activation memory or in parallel to suit different hardware constraints. By aggregating the predictions from these independent flow predictors, our framework enables robust, generative-inspired inference across diverse architectures, including CNNs and vision transformers.

Many machine learning classification tasks involve imbalanced datasets, which are often subject to over-sampling techniques aimed at improving model performance. However, these techniques are prone to generating unrealistic or infeasible samples. Furthermore, they often function as black boxes, lacking interpretability in their procedures. This opacity makes it difficult to track their effectiveness and provide necessary adjustments, and they may ultimately fail to yield significant performance improvements. To bridge this gap, we introduce the Decision Predicate Graphs for Data Augmentation (DPG-da), a framework that extracts interpretable decision predicates from trained models to capture domain rules and enforce them during sample generation. This design ensures that over-sampled data remain diverse, constraint-satisfying, and interpretable. In experiments on synthetic and real-world benchmark datasets, DPG-da consistently improves classification performance over traditional over-sampling methods, while guaranteeing logical validity and offering clear, interpretable explanations of the over-sampled data.

Vision-Language-Action (VLA) models have emerged as a powerful paradigm for robotic manipulation. However, existing post-training methods face a dilemma between stability and exploration: Supervised Fine-Tuning (SFT) is constrained by demonstration quality and lacks generalization, whereas Reinforcement Learning (RL) improves exploration but often induces erratic, jittery trajectories that violate physical constraints. To bridge this gap, we propose SmoothVLA, a novel reinforcement learning fine-tuning framework that synergistically optimizes task performance and motion smoothness. The technical core is a physics-informed hybrid reward function that integrates binary sparse task rewards with a continuous dense term derived from trajectory jerk. Crucially, this reward is intrinsic, that computing directly from policy rollouts, without requiring extrinsic environment feedback or laborious reward engineering. Leveraging the Group Relative Policy Optimization (GRPO), SmoothVLA establishes trajectory smoothness as an explicit optimization prior, guiding the model toward physically feasible and stable control. Extensive experiments on the LIBERO benchmark demonstrate that SmoothVLA outperforms standard RL by 13.8\% in smoothness and significantly surpasses SFT in generalization across diverse tasks. Our work offers a scalable approach to aligning VLA models with physical-world constraints through intrinsic reward optimization.

Cold metals are a class of metals with an intrinsic energy gap located close to the Fermi level, which enables cold-carrier injection for steep-slope transistors and is therefore promising for low-power electronic applications. High-throughput screening has revealed 252 three-dimensional (3D) cold metals in the Materials Project database, but database searches are inherently limited to known compounds. Here we present an inverse-design workflow that generates 3D cold metals using MatterGPT, a conditional autoregressive Transformer trained on SLICES, an invertible and symmetry-invariant crystal string representation. We curate a training set of 26,309 metallic structures labeled with energy above hull and a unified band-edge distance descriptor that merges p-type and n-type cold-metal characteristics to address severe label imbalance. Property-conditioned generation targeting thermodynamic stability and 50-500 meV band-edge distances produces 148,506 unique candidates; 92.1% are successfully reconstructed to 3D structures and down-selected by symmetry, uniqueness and novelty filters, followed by high-throughput DFT validation. We identify 257 cold metals verified as novel with respect to the Materials Project database, with gaps around the Fermi level spanning 50-500 meV. First-principles phonon, electronic-structure, and work-function calculations for representative candidates confirm dynamical stability and contact-relevant work functions. Our results demonstrate that SLICES-enabled generative transformers can expand the chemical space of cold metals beyond high-throughput screening, providing a route to low-power electronic materials discovery.

We show that language models hallucinate not because they fail to detect uncertainty, but because of a failure to integrate it into output generation. Across architectures, uncertain inputs are reliably identified, occupying high-dimensional regions with 2-3$\times$ the intrinsic dimensionality of factual inputs. However, this internal signal is weakly coupled to the output layer: uncertainty migrates into low-sensitivity subspaces, becoming geometrically amplified yet functionally silent. Topological analysis shows that uncertainty representations fragment rather than converging to a unified abstention state, while gradient and Fisher probes reveal collapsing sensitivity along the uncertainty direction. Because cross-entropy training provides no attractor for abstention and uniformly rewards confident prediction, associative mechanisms amplify these fractured activations until residual coupling forces a committed output despite internal detection. Causal interventions confirm this account by restoring refusal when uncertainty is directly connected to logits.

Federated learning (FL) enables a set of distributed clients to jointly train machine learning models while preserving their local data privacy, making it attractive for applications in healthcare, finance, mobility, and smart-city systems. However, FL faces several challenges, including statistical heterogeneity and uneven client participation, which can degrade convergence and model quality. In this work, we propose FedPBS, an FL algorithm that couples complementary ideas from FedBS and FedProx to address these challenges. FedPBS dynamically adapts batch sizes to client resources to support balanced and scalable participation, and selectively applies a proximal correction to small-batch clients to stabilize local updates and reduce divergence from the global model. Experiments on benchmarking datasets such as CIFAR-10 and UCI-HAR under highly non-IID settings demonstrate that FedPBS consistently outperforms state-of-the-art methods, including FedBS, FedGA, MOON, and FedProx. The results demonstrate robust performance gains under extreme data heterogeneity, with smooth loss curves indicating stable convergence across diverse federated environments. FedPBS consistently outperforms state-of-the-art federated learning baselines on UCI-HAR and CIFAR-10 under severe non-IID conditions while maintaining stable and reliable convergence.

For robotic agents operating in dynamic environments, learning visual state representations from streaming video observations is essential for sequential decision making. Recent self-supervised learning methods have shown strong transferability across vision tasks, but they do not explicitly address what a good visual state should encode. We argue that effective visual states must capture what-is-where by jointly encoding the semantic identities of scene elements and their spatial locations, enabling reliable detection of subtle dynamics across observations. To this end, we propose CroBo, a visual state representation learning framework based on a global-to-local reconstruction objective. Given a reference observation compressed into a compact bottleneck token, CroBo learns to reconstruct heavily masked patches in a local target crop from sparse visible cues, using the global bottleneck token as context. This learning objective encourages the bottleneck token to encode a fine-grained representation of scene-wide semantic entities, including their identities, spatial locations, and configurations. As a result, the learned visual states reveal how scene elements move and interact over time, supporting sequential decision making. We evaluate CroBo on diverse vision-based robot policy learning benchmarks, where it achieves state-of-the-art performance. Reconstruction analyses and perceptual straightness experiments further show that the learned representations preserve pixel-level scene composition and encode what-moves-where across observations.

Effective urban traffic monitoring is essential for improving mobility, enhancing safety, and supporting sustainable cities. Distributed Acoustic Sensing (DAS) enables large-scale traffic observation by transforming existing fiber-optic infrastructure into dense arrays of vibration sensors. However, modeling the high-resolution spatio-temporal structure of DAS data for reliable traffic event recognition remains challenging. This study presents a real-world DAS-based traffic monitoring experiment conducted in Granada, Spain, where vehicles cross a fiber deployed perpendicular to the roadway. Recurrent neural networks (RNNs) are employed to model intra- and inter-event temporal dependencies. Spatial and temporal attention mechanisms are systematically integrated within the RNN architecture to analyze their impact on recognition performance, parameter efficiency, and interpretability. Results show that an appropriate and complementary placement of attention modules improves the balance between accuracy and model complexity. Attention heatmaps provide physically meaningful interpretations of classification decisions by highlighting informative spatial locations and temporal segments. Furthermore, the proposed SA-bi-TA configuration demonstrates spatial transferability, successfully recognizing traffic events at sensing locations different from those used during training, with only moderate performance degradation. These findings support the development of scalable and interpretable DAS-based traffic monitoring systems capable of operating under heterogeneous urban sensing conditions.

Stacked AutoEncoders (SAE) have been widely adopted in edge anomaly detection scenarios. However, the resource-intensive nature of SAE can pose significant challenges for edge devices, which are typically resource-constrained and must adapt rapidly to dynamic and changing conditions. Optimizing SAE to meet the heterogeneous demands of real-world deployment scenarios, including high performance under constrained storage, low power consumption, fast inference, and efficient model updates, remains a substantial challenge. To address this, we propose an integrated optimization framework that jointly considers these critical factors to achieve balanced and adaptive system-level optimization. Specifically, we formulate SAE optimization for edge anomaly detection as a multi-objective optimization problem and propose MO-SAE (Multi-Objective Stacked AutoEncoders). The multiple objectives are addressed by integrating model clipping, multi-branch exit design, and a matrix approximation technique. In addition, a multi-objective heuristic algorithm is employed to effectively balance the competing objectives in SAE optimization. Our results demonstrate that the proposed MO-SAE delivers substantial improvements over the original approach. On the x86 architecture, it reduces storage space and power consumption by at least 50%, improves runtime efficiency by no less than 28%, and achieves an 11.8% compression rate, all while maintaining application performance. Furthermore, MO-SAE runs efficiently on edge devices with ARM architecture. Experimental results show a 15% improvement in inference speed, facilitating efficient deployment in cloud-edge collaborative anomaly detection systems.

Training deep neural networks with noisy labels remains a significant challenge, often leading to degraded performance. Existing methods for handling label noise typically rely on either transition matrix, noise detection, or meta-learning techniques, but they often exhibit low utilization efficiency of noisy samples and incur high computational costs. In this paper, we propose a self-training label correction framework using decoupled bilevel optimization, where a classifier and neural correction function co-evolve. Leveraging a small clean dataset, our method employs noisy posterior simulation and intermediate features to transfer ground-truth knowledge, forming a closed-loop feedback system that prevents error amplification. Theoretical guarantees underpin the stability of our approach, and extensive experiments on benchmark datasets like CIFAR and Clothing1M confirm state-of-the-art performance with reduced training time, highlighting its practical applicability for learning from noisy labels.

Vision-Language Models (VLMs) have emerged as powerful tools for image understanding tasks, yet their practical deployment remains hindered by significant architectural heterogeneity across model families. This paper introduces UVLM (Universal Vision-Language Model Loader), a Google Colab-based framework that provides a unified interface for loading, configuring, and benchmarking multiple VLM architectures on custom image analysis tasks. UVLM currently supports two major model families -- LLaVA-NeXT and Qwen2.5-VL -- which differ fundamentally in their vision encoding, tokenization, and decoding strategies. The framework abstracts these differences behind a single inference function, enabling researchers to compare models using identical prompts and evaluation protocols. Key features include a multi-task prompt builder with support for four response types (numeric, category, boolean, text), a consensus validation mechanism based on majority voting across repeated inferences, a flexible token budget (up to 1,500 tokens) enabling users to design custom reasoning strategies through prompt engineering, and a built-in chain-of-thought reference mode for benchmarking. UVLM is designed for reproducibility, accessibility, and extensibility and as such is freely deployable on Google Colab using consumer-grade GPU resources. The paper also presents the first benchmarking of different VLMs on tasks of increasing reasoning complexity using a corpus of 120 street-view images.

Large language models (LLMs) are increasingly used for automated text annotation in tasks ranging from academic research to content moderation and hiring. Across 19 LLMs and two experiments totaling more than 4 million annotation judgments, we show that subtle identity cues embedded in text systematically bias annotation outcomes in ways that mirror racial stereotypes. In a names-based experiment spanning 39 annotation tasks, texts containing names associated with Black individuals are rated as more aggressive by 18 of 19 models and more gossipy by 18 of 19. Asian names produce a bamboo-ceiling profile: 17 of 19 models rate individuals as more intelligent, while 18 of 19 rate them as less confident and less sociable. Arab names elicit cognitive elevation alongside interpersonal devaluation, and all four minority groups are consistently rated as less self-disciplined. In a matched dialect experiment, the same sentence is judged significantly less professional (all 19 models, mean gap $-0.774$), less indicative of an educated speaker ($-0.688$), more toxic (18/19), and more angry (19/19) when written in African American Vernacular English rather than Standard American English. A notable exception occurs for name-based hireability, where fine-tuning appears to overcorrect, systematically favoring minority-named applicants. These findings suggest that using LLMs as automated annotators can embed socially patterned biases directly into the datasets and measurements that increasingly underpin research, governance, and decision-making.

Large Language Models (LLMs) demonstrate significant potential for generating complex behaviors, yet most approaches lack mechanisms for modeling social motivation in human-like multi-agent interaction. We introduce Autonomous Social Value-Oriented agents (ASVO), where LLM-based agents integrate desire-driven autonomy with Social Value Orientation (SVO) theory. At each step, agents first update their beliefs by perceiving environmental changes and others' actions. These observations inform the value update process, where each agent updates multi-dimensional desire values through reflective reasoning and infers others' motivational states. By contrasting self-satisfaction derived from fulfilled desires against estimated others' satisfaction, agents dynamically compute their SVO along a spectrum from altruistic to competitive, which in turn guides activity selection to balance desire fulfillment with social alignment. Experiments across School, Workplace, and Family contexts demonstrate substantial improvements over baselines in behavioral naturalness and human-likeness. These findings show that structured desire systems and adaptive SVO drift enable realistic multi-agent social simulations.

Scene text recognition (STR) methods have demonstrated their excellent capability in English text images. However, due to the complex inner structures of Chinese and the extensive character categories, it poses challenges for recognizing Chinese text in images. Recently, studies have shown that the methods designed for English text recognition encounter an accuracy bottleneck when recognizing Chinese text images. This raises the question: Is it appropriate to apply the model developed for English to the Chinese STR task? To explore this issue, we propose a novel method named LER, which explicitly decouples each character and independently recognizes characters while taking into account the complex inner structures of Chinese. LER consists of three modules: Localization, Extraction, and Recognition. Firstly, the localization module utilizes multimodal information to determine the character's position precisely. Then, the extraction module dissociates all characters in parallel. Finally, the recognition module considers the unique inner structures of Chinese to provide the text prediction results. Extensive experiments conducted on large-scale Chinese benchmarks indicate that our method significantly outperforms existing methods. Furthermore, extensive experiments conducted on six English benchmarks and the Union14M benchmark show impressive results in English text recognition by LER. Code is available at https://github.com/Pandarenlql/LER.

Deploying trustworthy artificial intelligence on edge robotics imposes a difficult trade-off between high-assurance robustness and energy sustainability. Traditional defense mechanisms against adversarial attacks typically incur significant computational overhead, threatening the viability of power-constrained platforms in environments such as cislunar space. This paper quantifies the energy cost of assurance in event-driven neuromorphic systems. We benchmark the Hierarchical Temporal Defense (HTD) framework on the BrainChip Akida AKD1000 processor against a suite of adversarial temporal attacks. We demonstrate that unlike traditional deep learning defenses which often degrade efficiency significantly with increased robustness, the event-driven nature of the proposed architecture achieves a superior trade-off. The system reduces gradient-based adversarial success rates from 82.1% to 18.7% and temporal jitter success rates from 75.8% to 25.1%, while maintaining an energy consumption of approximately 45 microjoules per inference. We report a counter-intuitive reduction in dynamic power consumption in the fully defended configuration, attributed to volatility-gated plasticity mechanisms that induce higher network sparsity. These results provide empirical evidence that neuromorphic sparsity enables sustainable and high-assurance edge autonomy.

Chain-of-thought (CoT) reasoning has advanced medical visual question answering (VQA), yet most existing CoT rationales are free-form and fail to capture the structured reasoning process clinicians actually follow. This work asks: Can traceable, multi-step reasoning supervision improve reasoning accuracy and the interpretability of Medical VQA? To this end, we introduce Step-CoT, a large-scale medical reasoning dataset with expert-curated, structured multi-step CoT aligned to clinical diagnostic workflows, implicitly grounding the model's reasoning in radiographic evidence. Step-CoT comprises more than 10K real clinical cases and 70K VQA pairs organized around diagnostic workflows, providing supervised intermediate steps that guide models to follow valid reasoning trajectories. To effectively learn from Step-CoT, we further introduce a teacher-student framework with a dynamic graph-structured focusing mechanism that prioritizes diagnostically informative steps while filtering out less relevant contexts. Our experiments show that using Step-CoT can improve reasoning accuracy and interpretability. Benchmark: github.com/hahaha111111/Step-CoT. Dataset Card: huggingface.co/datasets/fl-15o/Step-CoT

The paper examines deep learning models for scribe verification in Chinese manuscripts. That is, to automatically determine whether two manuscript fragments were written by the same scribe using deep metric learning methods. Two datasets were used: the Tsinghua Bamboo Slips Dataset and a selected subset of the Multi-Attribute Chinese Calligraphy Dataset, focusing on the calligraphers with a large number of samples. Siamese and Triplet neural network architectures are implemented, including convolutional and Transformer-based models. The experimental results show that the MobileNetV3+ Custom Siamese model trained with contrastive loss achieves either the best or the second-best overall accuracy and area under the Receiver Operating Characteristic Curve on both datasets.

Do We Need Role Models? How do Role Models Shape Collective Morality? To explore the questions, we build a multi-agent simulation powered by a Large Language Model, where agents with diverse intrinsic drives, ranging from cooperative to competitive, interact and adapt through a four-stage cognitive loop (plan-act-observe-reflect). We design four experimental games (Alignment, Collapse, Conflict, and Construction) and conduct motivational ablation studies to identify the key drivers of imitation. The results indicate that identity-driven conformity can powerfully override initial dispositions. Agents consistently adapt their values to align with a perceived successful exemplar, leading to rapid value convergence.

Many large language model applications require conditioning on long contexts. Transformers typically support this by storing a large per-layer KV-cache of past activations, which incurs substantial memory overhead. A desirable alternative is ompressive memory: read a context once, store it in a compact state, and answer many queries from that state. We study this in a context removal setting, where the model must generate an answer without access to the original context at inference time. We introduce GradMem, which writes context into memory via per-sample test-time optimization. Given a context, GradMem performs a few steps of gradient descent on a small set of prefix memory tokens while keeping model weights frozen. GradMem explicitly optimizes a model-level self-supervised context reconstruction loss, resulting in a loss-driven write operation with iterative error correction, unlike forward-only methods. On associative key--value retrieval, GradMem outperforms forward-only memory writers with the same memory size, and additional gradient steps scale capacity much more effectively than repeated forward writes. We further show that GradMem transfers beyond synthetic benchmarks: with pretrained language models, it attains competitive results on natural language tasks including bAbI and SQuAD variants, relying only on information encoded in memory.

In 2020 Domingos introduced an interpolation formula valid for "every model trained by gradient descent". He concluded that such models behave approximately as kernel machines. In this work, we extend the Domingos formula to stochastic training. We introduce a stochastic gradient kernel that extends the deterministic version via a continuous-time diffusion approximation. We prove stochastic Domingos theorems and show that the expected network output admits a kernel-machine representation with optimizer-specific weighting. It reveals that training samples contribute through loss-dependent weights and gradient alignment along the training trajectory. We then link the generalization error to the null space of the integral operator induced by the stochastic gradient kernel. The same path-kernel viewpoint provides a unified interpretation of diffusion models and GANs: diffusion induces stage-wise, noise-localized corrections, whereas GANs induce distribution-guided corrections shaped by discriminator geometry. We visualize the evolution of implicit kernels during optimization and quantify out-of-distribution behaviors through a series of numerical experiments. Our results support a feature-space memory view of learning: training stores data-dependent information in an evolving tangent feature geometry, and predictions at test time arise from kernel-weighted retrieval and aggregation of these stored features, with generalization governed by alignment between test points and the learned feature memory.

Dexterous manipulation enables complex tasks but suffers from self-occlusion, severe depth noise, and depth information loss when manipulating transparent objects. To solve this problem, this paper proposes TransDex, a 3D visuo-tactile fusion motor policy based on point cloud reconstruction pre-training. Specifically, we first propose a self-supervised point cloud reconstruction pre-training approach based on Transformer. This method accurately recovers the 3D structure of objects from interactive point clouds of dexterous hands, even when random noise and large-scale masking are added. Building on this, TransDex is constructed in which perceptual encoding adopts a fine-grained hierarchical scheme and multi-round attention mechanisms adaptively fuse features of the robotic arm and dexterous hand to enable differentiated motion prediction. Results from transparent object manipulation experiments conducted on a real robotic system demonstrate that TransDex outperforms existing baseline methods. Further analysis validates the generalization capabilities of TransDex and the effectiveness of its individual components.

Building AI systems that can plan, act, and create in the physical world requires more than pattern recognition. Such systems must understand the causal mechanisms and constraints governing physical processes in order to guide sequential decisions. This capability relies on internal representations, analogous to an internal language model, that relate observations, actions, and resulting environmental changes. However, many existing benchmarks treat visual perception and programmatic reasoning as separate problems, focusing either on visual recognition or on symbolic tasks. The domain of origami provides a natural testbed that integrates these modalities. Constructing shapes through folding operations requires visual perception, reasoning about geometric and physical constraints, and sequential planning, while remaining sufficiently structured for systematic evaluation. We introduce OrigamiBench, an interactive benchmark in which models iteratively propose folds and receive feedback on physical validity and similarity to a target configuration. Experiments with modern vision-language models show that scaling model size alone does not reliably produce causal reasoning about physical transformations. Models fail to generate coherent multi-step folding strategies, suggesting that visual and language representations remain weakly integrated.

We introduce power term polynomial algebra, a representation language for Boolean formulae designed to bridge conjunctive normal form (CNF) and algebraic normal form (ANF). The language is motivated by the tiling mismatch between these representations: direct CNF<->ANF conversion may cause exponential blowup unless formulas are decomposed into smaller fragments, typically through auxiliary variables and side constraints. In contrast, our framework addresses this mismatch within the representation itself, compactly encoding structured families of monomials while representing CNF clauses directly, thereby avoiding auxiliary variables and constraints at the abstraction level. We formalize the language through power terms and power term polynomials, define their semantics, and show that they admit algebraic operations corresponding to Boolean polynomial addition and multiplication. We prove several key properties of the language: disjunctive clauses admit compact canonical representations; power terms support local shortening and expansion rewrite rules; and products of atomic terms can be systematically rewritten within the language. Together, these results yield a symbolic calculus that enables direct manipulation of formulas without expanding them into ordinary ANF. The resulting framework provides a new intermediate representation and rewriting calculus that bridges clause-based and algebraic reasoning and suggests new directions for structure-aware CNF<->ANF conversion and hybrid reasoning methods.

Retrieval-augmented generation (RAG), based on large language models (LLMs), serves as a vital approach to retrieving and leveraging external knowledge in various domain applications. When confronted with complex multi-hop questions, single-round retrieval is often insufficient for accurate reasoning and problem solving. To enhance search capabilities for complex tasks, most existing works integrate multi-round iterative retrieval with reasoning processes via end-to-end training. While these approaches significantly improve problem-solving performance, they are still faced with challenges in task reasoning and model training, especially ambiguous retrieval execution paths and sparse rewards in end-to-end reinforcement learning (RL) process, leading to inaccurate retrieval results and performance degradation. To address these issues, in this paper, we proposes APEX-Searcher, a novel Agentic Planning and Execution framework to augment LLM search capabilities. Specifically, we introduce a two-stage agentic framework that decouples the retrieval process into planning and execution: It first employs RL with decomposition-specific rewards to optimize strategic planning; Built on the sub-task decomposition, it then applies supervised fine-tuning on high-quality multi-hop trajectories to equip the model with robust iterative sub-task execution capabilities. Extensive experiments demonstrate that our proposed framework achieves significant improvements in both multi-hop RAG and task planning performances across multiple benchmarks.

Response variability limits the clinical utility of transcutaneous auricular vagus nerve stimulation (taVNS) for negative symptoms in treatment-resistant schizophrenia (TRS). This study aimed to develop an electroencephalography (EEG)-based machine learning (ML) model to predict individual response and explore associated neurophysiological mechanisms. We used ML to develop and validate predictive models based on pre-treatment EEG data features (power, coherence, and dynamic functional connectivity) from 50 TRS patients enrolled in the taVNS trial, within a nested cross-validation framework. Participants received 20 sessions of active or sham taVNS (n = 25 each) over two weeks, followed by a two-week follow-up. The prediction target was the percentage change in the positive and negative syndrome scale-factor score for negative symptoms (PANSS-FSNS) from baseline to post-treatment, with further evaluation of model specificity and neurophysiological relevance.The optimal model accurately predicted taVNS response in the active group, with predicted PANSS-FSNS changes strongly correlated with observed changes (r = 0.87, p < .001); permutation testing confirmed performance above chance (p < .001). Nine consistently retained features were identified, predominantly fronto-parietal and fronto-temporal coherence features. Negligible predictive performance in the sham group and failure to predict positive symptom change support the predictive specificity of this oscillatory signature for taVNS-related negative symptom improvement. Two coherence features within fronto-parietal-temporal networks showed post-taVNS changes significantly associated with symptom improvement, suggesting dual roles as predictors and potential therapeutic targets. EEG oscillatory neuromarkers enable accurate prediction of individual taVNS response in TRS, supporting mechanism-informed precision neuromodulation strategies.

We introduce EVE (Elemental Variational Expanse), a variational distributional neuron formulated as a local probabilistic computational unit with an explicit prior, an amortized posterior, and unit-level variational regularization. In most modern architectures, uncertainty is modeled through global latent variables or parameter uncertainty, while the computational unit itself remains scalar. EVE instead relocates probabilistic structure to the neuron level, making it locally observable and controllable. In this paper, the term dimensions refers primarily to the neuron's internal latent dimensionality, denoted by k. We study how varying k, from the atomic case k = 1 to higher-dimensional latent spaces, changes the neuron's learned operating regime. We then examine how this main axis interacts with two additional structural properties: local capacity control and temporal persistence through a neuron-level autoregressive extension. To support this study, EVE is instrumented with internal diagnostics and constraints, including effective KL, a target band on mu^2, out-of-band fractions, and indicators of drift and collapse. Across selected forecasting and tabular settings, we show that latent dimensionality, control, and temporal extension shape the neuron's internal regime, and that some neuron-level variables are measurable, informative, and related to downstream behavior. Overall, the paper provides an experimentally grounded first map of the design space opened by a variational neuron.

Speech-driven large language models (LLMs) are increasingly accessed through speech interfaces, introducing new security risks via open acoustic channels. We present Sirens' Whisper (SWhisper), the first practical framework for covert prompt-based attacks against speech-driven LLMs under realistic black-box conditions using commodity hardware. SWhisper enables robust, inaudible delivery of arbitrary target baseband audio-including long and structured prompts-on commodity devices by encoding it into near-ultrasound waveforms that demodulate faithfully after acoustic transmission and microphone nonlinearity. This is achieved through a simple yet effective approach to modeling nonlinear channel characteristics across devices and environments, combined with lightweight channel-inversion pre-compensation. Building on this high-fidelity covert channel, we design a voice-aware jailbreak generation method that ensures intelligibility, brevity, and transferability under speech-driven interfaces. Experiments across both commercial and open-source speech-driven LLMs demonstrate strong black-box effectiveness. On commercial models, SWhisper achieves up to 0.94 non-refusal (NR) and 0.925 specific-convincing (SC). A controlled user study further shows that the injected jailbreak audio is perceptually indistinguishable from background-only playback for human listeners. Although jailbreaks serve as a case study, the underlying covert acoustic channel enables a broader class of high-fidelity prompt-injection and commandexecution attacks.

Advances in machine learning have enabled the creation of realistic synthetic videos known as deepfakes. As deepfakes proliferate, concerns about rapid spread of disinformation and manipulation of public perception are mounting. Despite the alarming implications, our understanding of how individuals perceive synthetic media remains limited, obstructing the development of effective mitigation strategies. This paper aims to narrow this gap by investigating human responses to visual and auditory distortions of videos and deepfake-generated visuals and narration. In two between-subjects experiments, we study whether audio-visual distortions affect cognitive processing, such as subjective credibility assessment and objective learning outcomes. A third study reveals that artifacts from deepfakes influence credibility. The three studies show that video distortions and deepfake artifacts can reduce credibility. Our research contributes to the ongoing exploration of the cognitive processes involved in the evaluation and perception of synthetic videos, and underscores the need for further theory development concerning deepfake exposure.

End-to-end autonomous driving is typically built upon imitation learning (IL), yet its performance is constrained by the quality of human demonstrations. To overcome this limitation, recent methods incorporate reinforcement learning (RL) through sequential fine-tuning. However, such a paradigm remains suboptimal: sequential RL fine-tuning can introduce policy drift and often leads to a performance ceiling due to its dependence on the pretrained IL policy. To address these issues, we propose PaIR-Drive, a general Parallel framework for collaborative Imitation and Reinforcement learning in end-to-end autonomous driving. During training, PaIR-Drive separates IL and RL into two parallel branches with conflict-free training objectives, enabling fully collaborative optimization. This design eliminates the need to retrain RL when applying a new IL policy. During inference, RL leverages the IL policy to further optimize the final plan, allowing performance beyond prior knowledge of IL. Furthermore, we introduce a tree-structured trajectory neural sampler to group relative policy optimization (GRPO) in the RL branch, which enhances exploration capability. Extensive analysis on NAVSIMv1 and v2 benchmark demonstrates that PaIR-Drive achieves Competitive performance of 91.2 PDMS and 87.9 EPDMS, building upon Transfuser and DiffusionDrive IL baselines. PaIR-Drive consistently outperforms existing RL fine-tuning methods, and could even correct human experts' suboptimal behaviors. Qualitative results further confirm that PaIR-Drive can effectively explore and generate high-quality trajectories.

The complex interaction between social behaviors and climate change requires more than traditional data-driven prediction; it demands interpretable and adaptive analytical frameworks capable of integrating heterogeneous sources of knowledge. This study introduces ClimateAgents, a multi-agent research assistant designed to support social-climate analysis through coordinated AI agents. Rather than focusing solely on predictive modeling, the framework assists researchers in exploring socio-environmental dynamics by integrating multimodal data retrieval, statistical modeling, textual analysis, and automated reasoning. Traditional approaches to climate analysis often address narrowly defined indicators and lack the flexibility to incorporate cross-domain socio-economic knowledge or adapt to evolving research questions. To address these limitations, ClimateAgents employs a set of collaborative, domain-specialized agents that collectively perform key stages of the research workflow, including hypothesis generation, data analysis, evidence retrieval, and structured reporting. The framework supports exploratory analysis and scenario investigation using datasets from sources such as the United Nations and the World Bank. By combining agent-based reasoning with quantitative analysis of socio-economic behavioral dynamics, ClimateAgents enables adaptive and interpretable exploration of relationships between climate indicators, social variables, and environmental outcomes. The results illustrate how multi-agent AI systems can augment analytical reasoning and facilitate interdisciplinary, data-driven investigation of complex socio-environmental systems.

Modern data applications increasingly involve heterogeneous data managed in different models and stored across disparate database engines, often deployed as separate installs. Limited research has addressed cross-model query processing in federated environments. This paper takes a step toward bridging this gap by: (1) formally defining a class of cross-model join queries between a graph store and a relational store by proposing a unified algebra; (2) introducing one real-world benchmark and four semi-synthetic benchmarks to evaluate such queries; and (3) proposing a lightweight middleware, MICRO, for efficient query execution. At the core of MICRO is CMLero, a learning-to-rank-based query optimizer that selects efficient execution plans without requiring exact cost estimation. By avoiding the need to materialize or convert all data into a single model, which is often infeasible due to third-party data control or cost, MICRO enables native querying across heterogeneous systems. Experimental results on the benchmark workloads demonstrate that MICRO outperforms the state-of-the-art federated relational system XDB by up to 2.1x in total runtime across the full test set. On the 93 test queries of real-world benchmark, 14 queries achieve over 100 speedup, including 4 queries with more than 100x speedup; however, 4 queries experienced slowdowns of over 5 seconds, highlighting opportunities for future improvement of MICRO. Further comparisons show that CMLero consistently outperforms rule-based and regression-based optimizers, highlighting the advantage of learning-to-rank in complex cross-model optimization.

Predicting the mechanical properties of polysulfone (PSF) membranes from structural descriptors remains challenging due to extreme data scarcity typical of experimental studies. To investigate this issue, this study benchmarked knowledge-driven inference using four large language models (LLMs) (DeepSeek-V3, DeepSeek-R1, ChatGPT-4o, and GPT-5) against partial least squares (PLS) regression for predicting Young's modulus (E), tensile strength (TS), and elongation at break (EL) based on pore diameter (PD), contact angle (CA), thickness (T), and porosity (P) measurements. These knowledge-driven approaches demonstrated property-specific advantages over the chemometric baseline. For EL, LLMs achieved statistically significant improvements, with DeepSeek-R1 and GPT-5 delivering 40.5% and 40.3% of Root Mean Square Error reductions, respectively, reducing mean absolute errors from $11.63\pm5.34$% to $5.18\pm0.17$%. Run-to-run variability was markedly compressed for LLMs ($\leq$3%) compared to PLS (up to 47%). E and TS predictions showed statistical parity between approaches ($q\geq0.05$), indicating sufficient performance of linear methods for properties with strong structure-property correlations. Error topology analysis revealed systematic regression-to-the-mean behavior dominated by data-regime effects rather than model-family limitations. These findings establish that LLMs excel for non-linear, constraint-sensitive properties under bootstrap instability, while PLS remains competitive for linear relationships requiring interpretable latent-variable decompositions. The demonstrated complementarity suggests hybrid architectures leveraging LLM-encoded knowledge within interpretable frameworks may optimise small-data materials discovery.

Image segmentation is fundamental to microstructural analysis for defect identification and structure-property correlation, yet remains challenging due to pronounced heterogeneity in materials images arising from varied processing and testing conditions. Conventional image processing techniques often fail to capture such complex features rendering them ineffective for large-scale analysis. Even deep learning approaches struggle to generalize across heterogeneous datasets due to scarcity of high-quality labeled data. Consequently, segmentation workflows often rely on manual expert-driven annotations which are labor intensive and difficult to scale. Using an additive manufacturing (AM) dataset as a case study, we present a semi-automated active learning based segmentation pipeline that integrates a U-Net based convolutional neural network with an interactive user annotation and correction interface and a representative core-set image selection strategy. The active learning workflow iteratively updates the model by incorporating user corrected segmentations into the training pool while the core-set strategy identifies representative images for annotation. Three subset selection strategies, manual selection, uncertainty driven sampling and proposed maximin Latin hypercube sampling from embeddings (SMILE) method were evaluated over six refinement rounds. The SMILE strategy consistently outperformed other approaches, improving the macro F1 score from 0.74 to 0.93 while reducing manual annotation time by about 65 percent. The segmented defect regions were further analyzed using a coupled classification model to categorize defects based on microstructural characteristics and map them to corresponding AM process parameters. The proposed framework reduces labeling effort while maintaining scalability and robustness and is broadly applicable to image based analysis across diverse materials systems.

The high-frequency issuance and short-cycle speculation of meme tokens in decentralized finance (DeFi) have significantly amplified rug-pull risk. Existing approaches still struggle to provide stable early warning under scarce anomalies, incomplete labels, and limited interpretability. To address this issue, an end-to-end warning framework is proposed for BSC meme tokens, consisting of four stages: dataset construction and labeling, wash-trading pattern feature modeling, risk prediction, and error analysis. Methodologically, 12 token-level behavioral features are constructed based on three wash-trading patterns (Self, Matched, and Circular), unifying transaction-, address-, and flow-level signals into risk vectors. Supervised models are then employed to output warning scores and alert decisions. Under the current setting (7 tokens, 33,242 records), Random Forest outperforms Logistic Regression on core metrics, achieving AUC=0.9098, PR-AUC=0.9185, and F1=0.7429. Ablation results show that trade-level features are the primary performance driver (Delta PR-AUC=-0.1843 when removed), while address-level features provide stable complementary gain (Delta PR-AUC=-0.0573). The model also demonstrates actionable early-warning potential for a subset of samples, with a mean Lead Time (v1) of 3.8133 hours. The error profile (FP=1, FN=8) indicates that the current system is better positioned as a high-precision screener rather than a high-recall automatic alarm engine. The main contributions are threefold: an executable and reproducible rug-pull warning pipeline, empirical validation of multi-granularity wash-trading features under weak supervision, and deployment-oriented evidence through lead-time and error-bound analysis.

Human-computer interaction in the visual and auditory domains has achieved considerable maturity, yet machine-to-human tactile feedback remains underdeveloped. Existing tactile displays struggle to simultaneously render multiple tactile dimensions, such as shape, stiffness, and friction, which limits the realism of haptic simulation. Here, we present ArrayTac, a piezoelectric-driven tactile display capable of simultaneously rendering shape, stiffness, and friction to reproduce realistic haptic signals. The system comprises a 4x4 array of 16 actuator units, each employing a three-stage micro-lever mechanism to amplify the micrometer-scale displacement of the piezoelectric element, with Hall sensor-based closed-loop control at the end effector to enhance response speed and precision. We further implement two end-to-end pipelines: 1) a vision-to-touch framework that converts visual inputs into tactile signals using multimodal foundation models, and 2) a real-time tele-palpation system operating over distances of several thousand kilometers. In user studies, first-time participants accurately identify object shapes and physical properties with high success rates. In a tele-palpation experiment over 1,000km, untrained volunteers correctly identified both the number and type of tumors in a breast phantom with 100% accuracy and precisely localized their positions. The system pioneers a new pathway for high-fidelity haptic feedback by introducing the unprecedented capability to simultaneously render an object's shape, stiffness, and friction, delivering a holistic tactile experience that was previously unattainable.

This paper studies non-trivial consensus--a relatively novel and unexplored convergence behavior--on directed signed matrix-weighted networks subject to both additive and multiplicative measurement noises under time-varying topologies. Building upon grounded matrix-weighted Laplacian properties, a stochastic dynamic model is established that simultaneously captures inter-dimensional cooperative and antagonistic interactions, compound measurement noises and time-varying network structures. Based on stochastic differential equations theory, protocols that guarantee mean square and almost sure non-trivial consensus are proposed. Specifically, for any predetermined non-trivial consensus state, all agents are proven to converge toward this non-zero value in the mean-square and almost-sure senses. The design of control gain function in our protocols highlights a balanced consideration of the cumulative effect over time, the asymptotic decay property and the finite energy corresponding to measurement noises. Notably, the conditions on time-varying topologies in our protocols only require boundedness of elements in edge weight matrices, which facilitate the practicality of concept "time-varying topology" in matrix-weighted network consensus algorithms. Furthermore, the proposed protocols operate under milder connectivity conditions and no requirements on structural (un)balance properties. The work in this paper demonstrates that groups with both cooperative and antagonistic inter-dimensional interactions can achieve consensus even in the presence of compound measurement noises and time-varying topologies, challenging the conventional belief that consensus is attainable only in fully cooperative settings.

Classical sparsity promoting methods rely on the l0 norm, which treats all nonzero components as equally significant. In practical inverse problems, however, solutions often exhibit many small amplitude components that have little effect on reconstruction but lead to an overestimation of signal complexity. We address this limitation by shifting the paradigm from discrete cardinality to effective sparsity. Our approach introduces the effective number of nonzeros (ENZ), a unified class of normalized entropy-based regularizers, including Shannon and Renyi forms, that quantifies the concentration of significant coefficients. We show that, unlike the classical l0 norm, the ENZ provides a stable and continuous measure of effective sparsity that is insensitive to negligible perturbations. For noisy linear inverse problems, we establish theoretical guarantees under the Restricted Isometry Property (RIP), proving that ENZ based recovery is unique and stable. We also derive a decomposition showing that the ENZ equals the support cardinality times a distributional efficiency term, thereby linking entropy with l0 regularization. Numerical experiments show that this effective sparsity framework outperforms traditional cardinality based methods in robustness and accuracy.

Recent advances in text-to-audio generation enable models to translate natural-language descriptions into diverse musical output. However, the robustness of these systems under semantically equivalent prompt variations remains largely unexplored. Small linguistic changes may lead to substantial variation in generated audio, raising concerns about reliability in practical use. In this study, we evaluate the semantic fragility of text-to-audio systems under controlled prompt perturbations. We selected MusicGen-small, MusicGen-large, and Stable Audio 2.5 as representative models, and we evaluated them under Minimal Lexical Substitution (MLS), Intensity Shifts (IS), and Structural Rephrasing (SR). The proposed dataset contains 75 prompt groups designed to preserve semantic intent while introducing localized linguistic variation. Generated outputs are compared through complementary spectral, temporal, and semantic similarity measures, enabling robustness analysis across multiple representational levels. Experimental results show that larger models achieve improved semantic consistency, with MusicGen-large reaching cosine similarities of 0.77 under MLS and 0.82 under IS. However, acoustic and temporal analyses reveal persistent divergence across all models, even when embedding similarity remains high. These findings indicate that fragility arises primarily during semantic-to-acoustic realization rather than multi-modal embedding alignment. Our study introduces a controlled framework for evaluating robustness in text-to-audio generation and highlights the need for multi-level stability assessment in generative audio systems.

Precipitation nowcasting is vital for flood warning, agricultural management, and emergency response, yet two bottlenecks persist: the prohibitive cost of modeling million-scale spatiotemporal tokens from multi-variate atmospheric fields, and the extreme long-tailed rainfall distribution where heavy-to-torrential events -- those of greatest societal impact -- constitute fewer than 0.1% of all samples. We propose the Precipitation-Adaptive Network (PA-Net), a Transformer framework whose computational budget is explicitly governed by rainfall intensity. Its core component, Precipitation-Adaptive MoE (PA-MoE), dynamically scales the number of activated experts per token according to local precipitation magnitude, channeling richer representational capacity toward the rare yet critical heavy-rainfall tail. A Dual-Axis Compressed Latent Attention mechanism factorizes spatiotemporal attention with convolutional reduction to manage massive context lengths, while an intensity-aware training protocol progressively amplifies learning signals from extreme-rainfall samples. Experiment on ERA5 demonstrate consistent improvements over state-of-the-art baselines, with particularly significant gains in heavy-rain and rainstorm regimes.

Hospital logistics management faces growing pressure from internal operations and external emergencies, with artificial intelligence (AI) holding untapped potential to boost its resilience. This study explores AI's role in enhancing logistics resilience via a mixed-methods case study of H Hospital, combining 12 key informant interviews and a full survey of 151 logistics staff, with the PDCA cycle as the analytical framework. Thematic and quantitative analyses (hierarchical regression, structural equation modeling) were adopted for data analysis. Results showed 94.7% staff perceived AI application, with the strongest improvements in equipment maintenance (41.1%) and resource allocation (33.1%), but limited effects in emergency response (18.54%) and risk management (15.23%). AI integration positively correlated with logistics resilience (\b{eta}=0.642, p<0.001), with management system adaptability as a positive moderator (\b{eta}=0.208, p<0.01). The PDCA cycle fully mediated the AI-resilience relationship. We conclude AI effectively enhances logistics resilience, dependent on adaptive management systems and structured continuous improvement mechanisms. Targeted strategies are proposed to form an AI-driven closed-loop resilience mechanism, offering empirical guidance for AI-hospital logistics integration and resilient health system construction.

Spike sparsity is widely believed to enable efficient spiking neural network (SNN) inference on GPU hardware. We demonstrate this is an illusion: five distinct sparse computation strategies on Apple M3 Max all fail to outperform dense convolution, because SIMD architectures cannot exploit the fine-grained, unstructured sparsity of i.i.d. binary spikes. Instead, we propose Temporal Aggregated Convolution (TAC), which exploits convolution linearity to pre-aggregate $K$ spike frames before a single convolution call, reducing $T$ calls to $T/K$. On rate-coded data, TAC achieves 13.8times speedup with +1.6% accuracy on MNIST and +5.4% on Fashion-MNIST -- a simultaneous improvement in both speed and accuracy. However, on event-based data where the temporal dimension carries genuine motion information, TAC's temporal collapse is harmful. We therefore introduce TAC-TP (Temporal Preservation), which shares each group's convolution output across K independent LIF steps, preserving full temporal resolution for downstream layers. On DVS128-Gesture, TAC-TP achieves 95.1% accuracy (vs. 96.3% baseline) with 50% fewer convolution calls, while standard TAC drops to 91.3%. Our key finding is that the optimal temporal aggregation strategy is data-dependent: collapse the temporal dimension for rate-coded data (noise reduction) but preserve it for event data (information retention). Speedup is hardware-agnostic: TAC achieves 11.0times on NVIDIA V100, confirming the mechanism transfers across GPU architectures. All operators in the mlx-snn library are open source.

Sparse principal component analysis (SPCA) addresses the poor interpretability and variable redundancy often encountered by principal component analysis (PCA) in high-dimensional data. However, SPCA typically imposes uniform penalties on variables and does not account for differences in variable importance, which may lead to unstable performance in highly noisy or structurally complex settings. We propose SP-SPCA, a method that introduces a single equilibrium parameter into the regularization framework to adaptively adjust variable penalties. This modification of the L2 penalty provides flexible control over the trade-off between sparsity and explained variance while maintaining computational efficiency. Simulation studies show that the proposed method consistently outperforms standard sparse principal component methods in identifying sparse loading patterns, filtering noise variables, and preserving cumulative variance, especially in high-dimensional and noisy settings. Empirical applications to crime and financial market data further demonstrate its practical utility. In real data analyses, the method selects fewer but more relevant variables, thereby reducing model complexity while maintaining explanatory power. Overall, the proposed approach offers a robust and efficient alternative for sparse modeling in complex high-dimensional data, with clear advantages in stability, feature selection, and interpretability

Rehearsal-based continual learning (CL) mitigates catastrophic forgetting by maintaining a subset of samples from previous tasks for replay. Existing studies primarily focus on optimizing memory storage through coreset selection strategies. While these methods are effective, they typically require storing a substantial number of samples per class (SPC), often exceeding 20, to maintain satisfactory performance. In this work, we propose to further compress the memory footprint by synthesizing and storing prototypical exemplars, which can form representative prototypes when passed through the feature extractor. Owing to their representative nature, these exemplars enable the model to retain previous knowledge using only a small number of samples while preserving privacy. Moreover, we introduce a perturbation-based augmentation mechanism that generates synthetic variants of previous data during training, thereby enhancing CL performance. Extensive evaluations on widely used benchmark datasets and settings demonstrate that the proposed algorithm achieves superior performance compared to existing baselines, particularly in scenarios involving large-scale datasets and a high number of tasks.

Dynamic graph clustering aims to detect and track time-varying clusters in dynamic graphs, revealing how complex real-world systems evolve over time. However, existing methods are predominantly black-box models. They lack interpretability in their clustering decisions and fail to provide semantic explanations of why clusters form or how they evolve, severely limiting their use in safety-critical domains such as healthcare or transportation. To address these limitations, we propose an end-to-end interpretable framework that maps continuous graph embeddings into discrete semantic concepts through learnable prototypes. Specifically, we first decompose node representations into orthogonal role and clustering subspaces, so that nodes with similar roles (e.g., hubs, bridges) but different cluster affiliations can be properly distinguished. We then introduce five node role prototypes (Leader, Contributor, Wanderer, Connector, Newcomer) in the role subspace as semantic anchors, transforming continuous embeddings into discrete concepts to facilitate LLM understanding of node roles within communities. Finally, we design a hierarchical LLM reasoning mechanism to generate both clustering results and natural language explanations, while providing consistency feedback as weak supervision to refine node representations. Experimental results on four synthetic and six real-world benchmarks demonstrate the effectiveness, interpretability, and robustness of DyG-RoLLM. Code is available at https://github.com/Clearloveyuan/DyG-RoLLM.

High dialogue engagement is a crucial indicator of an effective conversation. A reliable measure of engagement could help benchmark large language models, enhance the effectiveness of human-computer interactions, or improve personal communication skills. However, quantifying engagement is challenging, since it is subjective and lacks a "gold standard". This paper proposes PMIScore, an efficient unsupervised approach to quantify dialogue engagement. It uses pointwise mutual information (PMI), which is the probability of generating a response conditioning on the conversation history. Thus, PMIScore offers a clear interpretation of engagement. As directly computing PMI is intractable due to the complexity of dialogues, PMIScore learned it through a dual form of divergence. The algorithm includes generating positive and negative dialogue pairs, extracting embeddings by large language models (LLMs), and training a small neural network using a mutual information loss function. We validated PMIScore on both synthetic and real-world datasets. Our results demonstrate the effectiveness of PMIScore in PMI estimation and the reasonableness of the PMI metric itself.

Federated Unlearning (FUL) aims to remove specific participants' data contributions from a trained Federated Learning model, thereby ensuring data privacy and compliance with regulatory requirements. Despite its potential, progress in FUL has been limited due to several challenges, including the cross-client knowledge inaccessibility and high computational and communication costs. To overcome these challenges, we propose Federated On-server Unlearning (FOUL), a novel framework that comprises two key stages. The learning-to-unlearn stage serves as a preparatory learning phase, during which the model identifies and encodes the key features associated with the forget clients. This stage is communication-efficient and establishes the basis for the subsequent unlearning process. Subsequently, on-server knowledge aggregation phase aims to perform the unlearning process at the server without requiring access to client data, thereby preserving both efficiency and privacy. We introduce a new data setting for FUL, which enables a more transparent and rigorous evaluation of unlearning. To highlight the effectiveness of our approach, we propose a novel evaluation metric termed time-to-forget, which measures how quickly the model achieves optimal unlearning performance. Extensive experiments conducted on three datasets under various unlearning scenarios demonstrate that FOUL outperforms the Retraining in FUL. Moreover, FOUL achieves competitive or superior results with significantly reduced time-to-forget, while maintaining low communication and computation costs.

Low resource languages present unique challenges for natural language processing due to the limited availability of digitized and well structured linguistic data. To address this gap, the GhanaNLP initiative has developed and curated 41,513 parallel sentence pairs for the Twi, Fante, Ewe, Ga, and Kusaal languages, which are widely spoken across Ghana yet remain underrepresented in digital spaces. Each dataset consists of carefully aligned sentence pairs between a local language and English. The data were collected, translated, and annotated by human professionals and enriched with standard structural metadata to ensure consistency and usability. These corpora are designed to support research, educational, and commercial applications, including machine translation, speech technologies, and language preservation. This paper documents the dataset creation methodology, structure, intended use cases, and evaluation, as well as their deployment in real world applications such as the Khaya AI translation engine. Overall, this work contributes to broader efforts to democratize AI by enabling inclusive and accessible language technologies for African languages.

As large language models (LLMs) scale to billions of parameters, full-parameter fine-tuning becomes compute- and memory-prohibitive. Parameter-efficient fine-tuning (PEFT) mitigates this issue by updating only a small set of task-specific parameters while keeping the base model frozen. Among PEFT approaches, low-rank adaptation (LoRA) is widely adopted; however, it enforces a uniform rank across layers despite substantial variation in layer importance, motivating {layerwise} rank allocation. Recent adaptive-rank variants (e.g., AdaLoRA) allocate ranks based on importance scores, yet typically rely on instantaneous gradients that capture only local sensitivity, overlooking non-local, pathwise effects within the same layer, which yields unstable and biased scores. To address this limitation, we introduce IGU-LoRA, an adaptive-rank LoRA that (i) computes within-layer Integrated Gradients (IG) sensitivities and aggregates them into a layer-level score for rank allocation, and (ii) applies an uncertainty-aware scheme using exponential moving averages with deviation tracking to suppress noisy updates and calibrate rank selection. Theoretically, we prove an upper bound on the composite trapezoidal rule approximation error for parameter-space IG under a pathwise Hessian-Lipschitz condition, which informs the quadrature budget. Across diverse tasks and architectures, IGU-LoRA consistently outperforms strong PEFT baselines at matched parameter budgets, improving downstream accuracy and robustness. Ablations confirm the contributions of pathwise within-layer sensitivity estimates and uncertainty-aware selection to effective rank allocation. Our code is publicly available at https://github.com/withyou12/igulora.git

Reliable detection of deceptive behavior in Large Language Model (LLM) agents is an essential prerequisite for safe deployment in high-stakes agentic contexts. Prior work on scheming detection has focused exclusively on black-box monitors that observe only externally visible tool calls and outputs, discarding potentially rich internal reasoning signals. We introduce DECEPTGUARD, a unified framework that systematically compares three monitoring regimes: black-box monitors (actions and outputs only), CoT-aware monitors (additionally observing the agent's chain-of-thought reasoning trace), and activation-probe monitors (additionally reading hidden-state representations from a frozen open-weights encoder). We introduce DECEPTSYNTH, a scalable synthetic pipeline for generating deception-positive and deception-negative agent trajectories across a novel 12-category taxonomy spanning verbal, behavioral, and structural deception. Our monitors are optimized on 4,800 synthetic trajectories and evaluated on 9,200 held-out samples from DeceptArena, a benchmark of realistic sandboxed agent environments with execution-verified labels. Across all evaluation settings, CoT-aware and activation-probe monitors substantially outperform their black-box counterparts (mean pAUROC improvement of +0.097), with the largest gains on subtle, long-horizon deception that leaves minimal behavioral footprints. We empirically characterize a transparency-detectability trade-off: as agents learn to suppress overt behavioral signals, chain-of-thought becomes the primary detection surface but is itself increasingly unreliable due to post-training faithfulness degradation. We propose HYBRID-CONSTITUTIONAL ensembles as a robust defense-in-depth approach, achieving a pAUROC of 0.934 on the held-out test set, representing a substantial advance over the prior state of the art.

We present \emph{Greedy Information Projection} (\textsc{GIP}), a principled framework for choosing training examples for large language model fine-tuning. \textsc{GIP} casts selection as maximizing mutual information between a subset of examples and task-specific query signals, which may originate from LLM quality judgments, metadata, or other sources. The framework involves optimizing a closed-form mutual information objective defined using both data and query embeddings, naturally balancing {\it quality} and {\it diversity}. Optimizing this score is equivalent to maximizing the projection of the query embedding matrix onto the span of the selected data, which provides a geometric explanation for the co-emergence of quality and diversity. Building on this view, we employ a fast greedy matching-pursuit procedure with efficient projection-based updates. On instruction-following and mathematical reasoning datasets, \textsc{GIP} selects small subsets that match full-data fine-tuning while using only a fraction of examples and compute, unifying quality-aware and diversity-aware selection for efficient fine-tuning.

Extreme-event predictability in turbulence is strongly state dependent, yet event-by-event predictability horizons are difficult to quantify without access to governing equations or costly perturbation ensembles. Here we train an autoregressive conditional diffusion model on direct numerical simulations of the two-dimensional Kolmogorov flow and use a CRPS-based skill score to define an event-wise predictability horizon. Enstrophy extremes exhibit a pronounced hierarchy: forecast skill persists from $\approx 1$ to $> 4$ Lyapunov times across events. Spectral filtering shows that these horizons are controlled predominantly by large-scale structures. Extremes are preceded by intense strain cores organizing quadrupolar vortex packets, whose lifetime sharply separates long- from short-horizon events. These results identify coherent-structure persistence as a governing mechanism for the predictability of turbulence extremes and provide a data-driven route to diagnose predictability limits from observations.

Continuous prompt search offers a computationally efficient alternative to conventional parameter tuning in natural language processing tasks. Nevertheless, its practical effectiveness can be significantly hindered by the black-box nature and the inherent high-dimensionality of the objective landscapes. Existing methods typically mitigate these challenges by restricting the search to a randomly projected low-dimensional subspace. However, the effectiveness and underlying motivation of the projection mechanism remain ambiguous. In this paper, we first empirically demonstrate that despite the prompt space possessing a low-dimensional structure, random projections fail to adequately capture this essential structure. Motivated by this finding, we propose a projection-free prompt search method based on evolutionary strategies. By directly optimizing in the full prompt space with an adaptation mechanism calibrated to the intrinsic dimension, our method achieves competitive search capabilities without additional computational overhead. Furthermore, to bridge the generalization gap in few-shot scenarios, we introduce a confidence-based regularization mechanism that systematically enhances the model's confidence in the target verbalizers. Experimental results on seven natural language understanding tasks from the GLUE benchmark demonstrate that our proposed approach significantly outperforms existing baselines.

Multimodal Large Language Models (MLLMs) have achieved impressive success in natural visual understanding, yet they consistently underperform in industrial anomaly detection (IAD). This is because MLLMs trained mostly on general web data differ significantly from industrial images. Moreover, they encode each image independently and can only compare images in the language space, making them insensitive to subtle visual differences that are key to IAD. To tackle these issues, we present AD-Copilot, an interactive MLLM specialized for IAD via visual in-context comparison. We first design a novel data curation pipeline to mine inspection knowledge from sparsely labeled industrial images and generate precise samples for captioning, VQA, and defect localization, yielding a large-scale multimodal dataset Chat-AD rich in semantic signals for IAD. On this foundation, AD-Copilot incorporates a novel Comparison Encoder that employs cross-attention between paired image features to enhance multi-image fine-grained perception, and is trained with a multi-stage strategy that incorporates domain knowledge and gradually enhances IAD skills. In addition, we introduce MMAD-BBox, an extended benchmark for anomaly localization with bounding-box-based evaluation. The experiments show that AD-Copilot achieves 82.3% accuracy on the MMAD benchmark, outperforming all other models without any data leakage. In the MMAD-BBox test, it achieves a maximum improvement of $3.35\times$ over the baseline. AD-Copilot also exhibits excellent generalization of its performance gains across other specialized and general-purpose benchmarks. Remarkably, AD-Copilot surpasses human expert-level performance on several IAD tasks, demonstrating its potential as a reliable assistant for real-world industrial inspection. All datasets and models will be released for the broader benefit of the community.

Aspect-based sentiment analysis (ABSA) extracts aspect-level sentiment signals from user-generated text, supports product analytics, experience monitoring, and public-opinion tracking, and is central to fine-grained opinion mining. A key challenge in ABSA is aspect sentiment quad prediction (ASQP), which requires identifying four elements: the aspect term, the aspect category, the opinion term, and the sentiment polarity. However, existing studies usually linearize the unordered quad set into a fixed-order template and decode it left-to-right. With teacher forcing training, the resulting training-inference mismatch (exposure bias) lets early prefix errors propagate to later elements. The linearization order determines which elements appear earlier in the prefix, so this propagation becomes order-sensitive and is hard to repair in a single pass. To address this, we propose a method, Generate-then-Correct (G2C): a generator drafts quads and a corrector performs a single-shot, sequence-level global correction trained on LLM-synthesized drafts with common error patterns. On the Rest15 and Rest16 datasets, G2C outperforms strong baseline models.

Large language models have recently enabled a generative paradigm for query expansion, but their high inference cost makes direct deployment difficult in practical retrieval systems. To address this issue, a retrieval-feedback-driven distillation and preference-alignment framework is proposed to transfer retrieval-friendly expansion behavior from a strong teacher model to a compact student model. Rather than relying on few-shot exemplars at inference time, the framework first leverages two complementary types of teacher-generated expansions, produced under zero-shot and few-shot prompting conditions, as supervision signals for distillation and as candidate pools for preference construction. A retrieval-metric-driven strategy is then introduced to automatically form chosen/rejected expansion pairs according to nDCG@10 differences, and Direct Preference Optimization is applied to explicitly align generation preferences with retrieval objectives. Experiments on TREC DL19/20/21 and MIRACL-zh show that the proposed approach preserves strong retrieval effectiveness while substantially reducing inference cost. In particular, the distilled Qwen3-4B model reaches about 97% of the teacher (DeepSeek-685B) model's nDCG@10 performance on DL19, and remains effective on the Chinese MIRACL-zh benchmark, demonstrating strong practicality across both English and Chinese retrieval settings.

Web information extraction (WIE) is the task of automatically extracting data from web pages, offering high utility for various applications. The evaluation of WIE systems has traditionally relied on benchmarks built from HTML snapshots captured at a single point in time. However, this offline evaluation paradigm fails to account for the temporally evolving nature of the web; consequently, performance on these static benchmarks often fails to generalize to dynamic real-world scenarios. To bridge this gap, we introduce \dataset, a new benchmark designed for evaluating WIE systems directly against live websites. Based on trusted and permission-granted websites, we curate natural language queries that require information extraction of various data categories, such as text, images, and hyperlinks. We further design these queries to represent four levels of complexity, based on the number and cardinality of attributes to be extracted, enabling a granular assessment of WIE systems. In addition, we propose Visual Grounding Scraper (VGS), a novel multi-stage agentic framework that mimics human cognitive processes by visually narrowing down web page content to extract desired information. Extensive experiments across diverse backbone models demonstrate the effectiveness and robustness of VGS. We believe that this study lays the foundation for developing practical and robust WIE systems.

Accurate and interpretable brain tumor classification from medical imaging remains a challenging problem due to the high dimensionality and complex structural patterns present in magnetic resonance imaging (MRI). In this study, we propose a topology-driven framework for brain tumor classification based on Topological Data Analysis (TDA) applied directly to three-dimensional (3D) MRI volumes. Specifically, we analyze 3D Fluid Attenuated Inversion Recovery (FLAIR) images from the BraTS 2020 dataset and extract interpretable topological descriptors using persistent homology. Persistent homology captures intrinsic geometric and structural characteristics of the data through Betti numbers, which describe connected components (Betti-0), loops (Betti-1), and voids (Betti-2). From the 3D MRI volumes, we derive a compact set of 100 topological features that summarize the underlying topology of brain tumor structures. These descriptors represent complex 3D tumor morphology while significantly reducing data dimensionality. Unlike many deep learning approaches that require large-scale training data or complex architectures, the proposed framework relies on computationally efficient topological features extracted directly from the images. These features are used to train classical machine learning classifiers, including Random Forest and XGBoost, for binary classification of high-grade glioma (HGG) and low-grade glioma (LGG). Experimental results on the BraTS 2020 dataset show that the Random Forest classifier combined with selected Betti features achieves an accuracy of 89.19%. These findings highlight the potential of persistent homology as an effective and interpretable approach for analyzing complex 3D medical images and performing brain tumor classification.

Despite the strong performance of large audio language models (LALMs) in various tasks, exactly how and where they integrate acoustic features with textual context remains unclear. We adapt causal tracing to investigate the internal information flow of LALMs during audio comprehension. By conducting layer-wise and token-wise analyses across DeSTA, Qwen, and Voxtral, we evaluate the causal effects of individual hidden states. Layer-wise analysis identifies different fusion strategies, from progressive integration in DeSTA to abrupt late-stage fusion in Qwen. Token-wise analysis shows that the final sequence token acts as an informational bottleneck where the network decisively retrieves relevant information from the audio. We also observe an attention-like query mechanism at intermediate token positions that triggers the model to pull task-relevant audio context. These findings provide a clear characterization of when and where multi-modal integration occurs within LALMs.

Modern chip design requires multi-objective optimization of timing, power, and area under stringent time-to-market constraints. Although powerful optimization algorithms are integrated into EDA tools, achieving high QoR hinges on effective long-horizon scheduling, which relies heavily on manual expert intervention. To address this issue and automate chip design, we propose an agentic LLM framework that schedules chip optimizations through direct interaction with EDA tools. The agent is grounded in natural language expertise expressed as a search tree through retrieval-augmented generation (RAG). We further improve scheduling quality with Pareto-driven QoR feedback through language reflection. Experimental results show that, compared with black-box search methods such as reinforcement learning, our framework achieves 10% greater timing improvement while consuming less power and area, with more than 4x speedup. The post-optimization QoR is also comparable to that achieved by human experts. Finally, the agent supports customized tasks expressed in natural language, enabling preferential QoR trade-offs. The code and chip design data will be publicly available at https://github.com/YiKangOY/Open-LLM-ECO.

We propose a resource-efficient framework for compressing large language models through knowledge distillation, combined with guided chain-of-thought reinforcement learning. Using Qwen 3B as the teacher and Qwen 0.5B as the student, we apply knowledge distillation across English Dolly-15k, Spanish Dolly-15k, and code BugNet and PyTorrent datasets, with hyperparameters tuned in the English setting to optimize student performance. Across tasks, the distilled student retains a substantial portion of the teacher's capability while remaining significantly smaller: 70% to 91% in English, up to 95% in Spanish, and up to 93.5% Rouge-L in code. For coding tasks, integrating chain-of-thought prompting with Group Relative Policy Optimization using CoT-annotated Codeforces data improves reasoning coherence and solution correctness compared to knowledge distillation alone. Post-training 4-bit weight quantization further reduces memory footprint and inference latency. These results show that knowledge distillation combined with chain-of-thought guided reinforcement learning can produce compact, efficient models suitable for deployment in resource-constrained settings.

Curriculum learning--ordering training examples in a sequence to aid machine learning--takes inspiration from human learning, but has not gained widespread acceptance. Static strategies for scoring item difficulty rely on indirect proxy scores of varying quality and produce curricula that are not specific to the learner at hand. Dynamic approaches base difficulty estimates on gradient information, requiring considerable extra computation during training. We introduce a novel method for measuring the difficulty of individual problem instances directly relative to the ability of a given model, and identify transitional problems that are consistently easier as model ability increases. Applying this method to chess and mathematics, we find that training on a curriculum that "levels up" from easier to harder transitional problems most efficiently improves a model to the next tier of competence. These problems induce a natural progression from easier to harder items, which outperforms other training strategies. By measuring difficulty directly relative to model competence, our method yields interpretable problems, learner-specific curricula, and a principled basis for step-by-step improvement.

We participated in the 10th ABAW Challenge, focusing on the Emotional Mimicry Intensity (EMI) Estimation track on the Hume-Vidmimic2 dataset. This task aims to predict six continuous emotion dimensions: Admiration, Amusement, Determination, Empathic Pain, Excitement, and Joy. Through systematic multimodal exploration of pretrained high-level features, we found that, under our pretrained feature setting, direct feature concatenation outperformed the more complex fusion strategies we tested. This empirical finding motivated us to design a systematic approach built upon three core principles: (i) preserving modality-specific attributes through feature-level concatenation; (ii) improving training stability and metric alignment via multi-objective optimization; and (iii) enriching acoustic representations with a VAD-inspired latent prior. Our final framework integrates concatenation-based multimodal fusion, a shared six-dimensional regression head, multi-objective optimization with MSE, Pearson-correlation, and auxiliary branch supervision, EMA for parameter stabilization, and a VAD-inspired latent prior for the acoustic branch. On the official validation set, the proposed scheme achieved our best mean Pearson Correlation Coefficient of 0.478567.

Recently, Transformer-based architectures have advanced meteorological prediction. However, this position-centric tokenizer conflicts with the core principle of meteorological systems, where the weather phenomena undoubtedly involve synergistic interactions among multiple elements while positional information constitutes merely a component of the boundary conditions. This paper focuses primarily on the task of precipitation nowcasting and develops an efficient distribution-centric Meteorological Tokenization (MeTok) scheme, which spatially sequences to group similar meteorological features. Based on the rearrangement, realigned group learning enhances robustness across precipitation patterns, especially extreme ones. Specifically, we introduce the Hyper-Aligned Grouping Transformer (HyAGTransformer) with two key improvements: 1) The Grouping Attention (GA) mechanism uses MeTok to enable self-aligned learning of features from different precipitation patterns; 2) The Neighborhood Feed-Forward Network (N-FFN) integrates adjacent group features, aggregating contextual information to boost patch embedding discriminability. Experiments on the ERA5 dataset for 6-hour forecasts show our method improves the IoU metric by at least 8.2% in extreme precipitation prediction compared to other methods. Additionally, it gains performance with more training data and increased parameters, demonstrating scalability, stability, and superiority over traditional methods.

Physics-informed neural networks (PINNs) have achieved notable success in modeling dynamical systems governed by partial differential equations (PDEs). To avoid computationally expensive retraining under new physical conditions, parameterized PINNs (P$^2$INNs) commonly adapt pre-trained operators using singular value decomposition (SVD) for out-of-distribution (OOD) regimes. However, SVD-based fine-tuning often suffers from rigid subspace locking and truncation of important high-frequency spectral modes, limiting its ability to capture complex physical transitions. While parameter-efficient fine-tuning (PEFT) methods appear to be promising alternatives, applying conventional adapters such as LoRA to P$^2$INNs introduces a severe Pareto trade-off, as additive updates increase parameter overhead and disrupt the structured physical manifolds inherent in operator representations. To address these limitations, we propose Manifold-Orthogonal Dual-spectrum Extrapolation (MODE), a lightweight micro-architecture designed for physics operator adaptation. MODE decomposes physical evolution into complementary mechanisms including principal-spectrum dense mixing that enables cross-modal energy transfer within frozen orthogonal bases, residual-spectrum awakening that activates high-frequency spectral components through a single trainable scalar, and affine Galilean unlocking that explicitly isolates spatial translation dynamics. Experiments on challenging PDE benchmarks including the 1D Convection--Diffusion--Reaction equation and the 2D Helmholtz equation demonstrate that MODE achieves strong out-of-distribution generalization while preserving the minimal parameter complexity of native SVD and outperforming existing PEFT-based baselines.

The datacenter industry is converging on SmartNIC-based resource management. Wave (Humphries et al., ASPLOS '25) demonstrates the practical feasibility of offloading kernel thread scheduling, memory management, and RPC stacks to the ARM cores of Intel's Mount Evans Infrastructure Processing Unit (IPU). The engineering is careful and the results are honest: without Wave's PCIe latency mitigations, offloaded workloads degrade by 350%. We argue that this 350% degradation is not an engineering problem to be optimized away but a diagnostic symptom of a deeper architectural issue: Wave's communication model is Forward-In-Time-Only (FITO). Every interaction between host and SmartNIC is a unidirectional message -- event forward, decision back -- creating a temporal vulnerability window in which decisions can become stale before they are enforced. Wave's entire optimization stack (write-combining page table entries, prestaging, prefetching, atomic transaction abort) exists to hide or tolerate this window. We apply the FITO diagnostic to Wave's architecture systematically, identify the category mistake it inherits from Lamport's happened-before and Shannon's channel model, and show how Open Atomic Ethernet's bilateral swap primitive -- implemented on the same Intel IPU hardware -- dissolves the latency, atomicity, and timeout problems without engineering around them. The SmartNIC is the right location for resource management; what is missing is the right communication primitive at that location.

Detecting subtle deviations in noisy acoustic environments is central to anomalous sound detection (ASD). A common training-free ASD pipeline temporally pools frame-level representations into a band-preserving feature vector and scores anomalies using a single nearest-neighbor match. However, this global matching can inflate normal-score variance through two effects. First, when normal sounds exhibit band-wise variability, a single global neighbor forces all bands to share the same reference, increasing band-level mismatch. Second, cosine-based matching is energy-coupled, allowing a few high-energy bands to dominate score computation under normal energy fluctuations and further increase variance. We propose BEAM, which stores temporally pooled sub-band vectors in a memory bank, retrieves neighbors per sub-band, and uniformly aggregates scores to reduce normal-score variability and improve discriminability. We further introduce a parameter-free adaptive fusion to better handle diverse temporal dynamics in sub-band responses. Experiments on multiple DCASE Task 2 benchmarks show strong performance without task-specific training, robustness to noise and domain shifts, and complementary gains when combined with encoder fine-tuning.

Real-world robots often operate in settings where objective priorities depend on the underlying context of operation. When the underlying context is unknown apriori, multiple robots may have to coordinate to gather informative observations to infer the context, since acting based on an incorrect context can lead to misaligned and unsafe behavior. Once the underlying true context is inferred, the robots optimize their task-specific objectives in the preference order induced by the context. We formalize this problem as a Multi-Robot Context-Uncertain Stochastic Shortest Path (MR-CUSSP), which captures context-relevant information at landmark states through joint observations. Our two-stage solution approach is composed of: (1) CIMOP (Coordinated Inference for Multi-Objective Planning) to compute plans that guide robots toward informative landmarks to efficiently infer the true context, and (2) LCBS (Lexicographic Conflict-Based Search) for collision-free multi-robot path planning with lexicographic objective preferences, induced by the context. We evaluate the algorithms using three simulated domains and demonstrate its practical applicability using five mobile robots in the salp domain setup.

The classical material tetrahedron that represents the Structure-Property-Processing-Performance-Characterization relationship is the most important research paradigm in materials science so far. It has served as a protocol to guide experiments, modeling, and theory to uncover hidden relationships between various aspects of a certain material. This substantially facilitates knowledge accumulation and material discovery with desired functionalities to realize versatile applications. In recent years, with the advent of artificial intelligence (AI) techniques, the attention of AI towards scientific research is soaring. The trials of implementing AI in various disciplines are endless, with great potential to revolutionize the research diagram. Despite the success in natural language processing and computer vision, how to effectively integrate AI with natural science is still a grand challenge, bearing in mind their fundamental differences. Inspired by these observations and limitations, we delve into the current research paradigm dictated by the classical material tetrahedron and propose two new paradigms to stimulate data-driven and AI-augmented research. One tetrahedron focuses on AI for materials science by considering the Matter-Data-Model-Potential-Agent diagram. The other demonstrates AI research by discussing Data-Architecture-Encoding-Optimization-Inference relationships. The crucial ingredients of these frameworks and their connections are discussed, which will likely motivate both scientific thinking refinement and technology advancement. Despite the widespread enthusiasm for chasing AI for science, we must analyze issues rationally to come up with well-defined, resolvable scientific problems in order to better master the power of AI.

We study stochastic multi-armed bandits under simultaneous constraints on space and adaptivity: the learner interacts with the environment in $B$ batches and has only $W$ bits of persistent memory. Prior work shows that each constraint alone is surprisingly mild: near-minimax regret $\widetilde{O}(\sqrt{KT})$ is achievable with $O(\log T)$ bits of memory under fully adaptive interaction, and with a $K$-independent $O(\log\log T)$-type number of batches when memory is unrestricted. We show that this picture breaks down in the simultaneously constrained regime. We prove that any algorithm with a $W$-bit memory constraint must use at least $Ω(K/W)$ batches to achieve near-minimax regret $\widetilde{O}(\sqrt{KT})$ , even under adaptive grids. In particular, logarithmic memory rules out $K$-independent batch complexity. Our proof is based on an information bottleneck. We show that near-minimax regret forces the learner to acquire $Ω(K)$ bits of information about the hidden set of good arms under a suitable hard prior, whereas an algorithm with $B$ batches and $W$ bits of memory allows only $O(BW)$ bits of information. A key ingredient is a localized change-of-measure lemma that yields probability-level arm exploration guarantees, which is of independent interest. We also give an algorithm using $O(\log T)$ bits of memory and $\widetilde{O}(K)$ batches that achieves regret $\widetilde{O}(\sqrt{KT})$, which nearly matches our lower bound.

Diffusion-based text-to-video generation (T2V) or image-to-video (I2V) generation have emerged as a prominent research focus. However, there exists a challenge in integrating the two generative paradigms into a unified model. In this paper, we present a unified video generation model (UniVid) with hybrid conditions of the text prompt and reference image. Given these two available controls, our model can extract objects' appearance and their motion descriptions from textual prompts, while obtaining texture details and structural information from image clues to guide the video generation process. Specifically, we scale up the pre-trained text-to-image diffusion model for generating temporally coherent frames via introducing our temporal-pyramid cross-frame spatial-temporal attention modules and convolutions. To support bimodal control, we introduce a dual-stream cross-attention mechanism, whose attention scores can be freely re-weighted for interpolation of between single and two modalities controls during inference. Extensive experiments showcase that our UniVid achieves superior temporal coherence on T2V, I2V and (T+I)2V tasks.

The Markov approximation is arguably the most ubiquitous tool in physics, underpinning quantum master equations, stochastic processes, and -- via Shannon's channel model and Lamport's logical clocks -- the foundational assumptions of distributed computing. It is widely assumed that Markovianity inherently implies temporal asymmetry: that the Markov property is a forward-in-time-only (FITO) construct. We show that this assumption is a category mistake in the sense of Ryle (1949). Guff, Shastry, and Rocco (2025) have recently demonstrated that the Markov approximation applied to the Caldeira-Leggett model -- a paradigmatic open quantum system -- maintains time-reversal symmetry in the derived equations of motion. The resulting time-symmetric formulations of quantum Brownian motion, Lindblad master equations, and Pauli master equations describe thermalisation that can occur in two opposing temporal directions. Asymmetry arises not from the dynamics but from boundary conditions. We trace how Markovianity's assumed directionality propagated from physics through Shannon's information theory to Lamport's happens-before relation and the impossibility theorems of distributed computing (FLP, CAP, Two Generals). Each step encodes FITO as convention, then treats it as physical law -- the same category mistake repeated across domains. The Surrey result establishes that this conflation is not merely philosophically suspect but mathematically unnecessary: the most fundamental approximation used to derive irreversibility is itself time-symmetric.

Diffusion-based models have recently shown strong performance in trajectory planning, as they are capable of capturing diverse, multimodal distributions of complex behaviors. A key limitation of these models is their slow inference speed, which results from the iterative denoising process. This makes them less suitable for real-time applications such as closed-loop model predictive control (MPC), where plans must be generated quickly and adapted continuously to a changing environment. In this paper, we investigate Implicit Maximum Likelihood Estimation (IMLE) as an alternative generative modeling approach for planning. IMLE offers strong mode coverage while enabling inference that is two orders of magnitude faster, making it particularly well suited for real-time MPC tasks. Our results demonstrate that IMLE achieves competitive performance on standard offline reinforcement learning benchmarks compared to the standard diffusion-based planner, while substantially improving planning speed in both open-loop and closed-loop settings. We further validate IMLE in a closed-loop human navigation scenario, operating in real-time, demonstrating how it enables rapid and adaptive plan generation in dynamic environments.

This paper addresses two persistent challenges in sequential recommendation: (i) evidence insufficiency-cold-start sparsity together with noisy, length-varying item texts; and (ii) opaque modeling of dynamic, multi-faceted intents across long/short horizons. We propose R3-REC (Reasoning-Retrieval-Recommendation), a prompt-centric, retrieval-augmented framework that unifies Multi-level User Intent Reasoning, Item Semantic Extraction, Long-Short Interest Polarity Mining, Similar User Collaborative Enhancement, and Reasoning-based Interest Matching and Scoring. Across ML-1M, Games, and Bundle, R3-REC consistently surpasses strong neural and LLM baselines, yielding improvements up to +10.2% (HR@1) and +6.4% (HR@5) with manageable end-to-end latency. Ablations corroborate complementary gains of all modules.

Symbolic regression is a powerful tool for knowledge discovery, enabling the extraction of interpretable mathematical expressions directly from data. However, conventional symbolic discovery typically follows an end-to-end, "one-step" process, which often generates lengthy and physically meaningless expressions when dealing with real physical systems, leading to poor model generalization. This limitation fundamentally stems from its deviation from the basic path of scientific discovery: physical laws do not exist in a single form but follow a hierarchical and progressive pattern from simplicity to complexity. Motivated by this principle, we propose Chain of Symbolic Regression (CoSR), a novel framework that models the discovery of physical laws as a chain of symbolic knowledge. This knowledge chain is formed by progressively combining multiple knowledge units with clear physical meanings along a specific logic, ultimately enabling the precise discovery of the underlying physical laws from data. CoSR fully recapitulates the progressive discovery path from Kepler's third law to the law of universal gravitation in classical mechanics, and is applied to three types of problems: turbulent Rayleigh-Benard convection, viscous flows in a circular pipe, and laser-metal interaction, demonstrating its ability to improve classical scaling theories. Finally, CoSR showcases its capability to discover new knowledge in the complex engineering problem of aerodynamic coefficients scaling for different aircraft.

Memristor-based analog compute-in-memory (CIM) architectures provide a promising substrate for the efficient deployment of Large Language Models (LLMs), owing to superior energy efficiency and computational density. However, these architectures suffer from precision issues caused by intrinsic non-idealities of memristors. In this paper, we first conduct a comprehensive investigation into the impact of such typical non-idealities on LLM reasoning. Empirical results indicate that reasoning capability decreases significantly but varies for distinct benchmarks. Subsequently, we systematically appraise three training-free strategies, including thinking mode, in-context learning, and module redundancy. We thus summarize valuable guidelines, i.e., shallow layer redundancy is particularly effective for improving robustness, thinking mode performs better under low noise levels but degrades at higher noise, and in-context learning reduces output length with a slight performance trade-off. Our findings offer new insights into LLM reasoning under non-ideality and practical strategies to improve robustness.

Agent-based coding tools have transformed software development practices. Unlike prompt-based approaches that require developers to manually integrate generated code, these agent-based tools autonomously interact with repositories to create, modify, and execute code, including test generation. While many developers have adopted agent-based coding tools, little is known about how these tools generate tests in real-world development scenarios or how AI-generated tests compare to human-written ones. This study presents an empirical analysis of test generation by agent-based coding tools using the AIDev dataset. We extracted 2,232 commits containing test-related changes and investigated three aspects: the frequency of test additions, the structural characteristics of the generated tests, and their impact on code coverage. Our findings reveal that (i) AI authored 16.4% of all commits adding tests in real-world repositories, (ii) AI-generated test methods exhibit distinct structural patterns, featuring longer code and a higher density of assertions while maintaining lower cyclomatic complexity through linear logic, and (iii) AI-generated tests contribute to code coverage comparable to human-written tests, frequently achieving positive coverage gains across several projects.

Code readability is fundamental to software quality and maintainability. Poor readability extends development time, increases bug-inducing risks, and contributes to technical debt. With the rapid advancement of Large Language Models, AI agent-based approaches have emerged as a promising paradigm for automated refactoring, capable of decomposing complex tasks through autonomous planning and execution. While prior studies have examined refactoring by AI agents, these analyses cover all forms of refactoring, including performance optimization and structural improvement. As a result, the extent to which AI agent-based refactoring specifically improves code readability remains unclear. This study investigates the impact of AI agent-based refactoring on code readability. We extracted commits containing readability-related keywords from the AIDev dataset and analyzed changes in readability metrics before and after each commit, covering 403 commits evaluated using multiple quantitative metrics. Our results indicate that AI agents primarily target logic complexity (42.4%) and documentation improvements (24.2%) rather than surface-level aspects like naming conventions or formatting. However, contrary to expectations, readability-focused commits often degraded traditional quality metrics: the Maintainability Index decreased in 56.1% of commits, while Cyclomatic Complexity increased in 42.7%.

Home-based interventions like parent-child shared reading provide a cost-effective approach for supporting children with autism spectrum disorder (ASD). However, analyzing caregiver intervention strategies in naturalistic home interactions typically relies on expert annotation, which is costly, time-intensive, and difficult to scale. To address this challenge, we propose InterventionLens, an end-to-end multi-agent system for automatically detecting and temporally segmenting caregiver intervention strategies from shared reading videos. Without task-specific model training or fine-tuning, InterventionLens uses a collaborative multi-agent architecture to integrate multimodal interaction content and perform fine-grained strategy analysis. Experiments on the ASD-HI dataset show that InterventionLens achieves an overall F1 score of 79.44\%, outperforming the baseline by 19.72\%. These results suggest that InterventionLens is a promising system for analyzing caregiver intervention strategies in home-based ASD shared reading settings. Additional resources will be released on the project page.

Humanoid loco-manipulation requires coordinated high-level motion plans with stable, low-level whole-body execution under complex robot-environment dynamics and long-horizon tasks. While diffusion policies (DPs) show promise for learning from demonstrations, deploying them on humanoids poses critical challenges: the motion planner trained offline is decoupled from the low-level controller, leading to poor command tracking, compounding distribution shift, and task failures. The common approach of scaling demonstration data is prohibitively expensive for high-dimensional humanoid systems. To address this challenge, we present REFINE-DP (REinforcement learning FINE-tuning of Diffusion Policy), a hierarchical framework that jointly optimizes a DP high-level planner and an RL-based low-level loco-manipulation controller. The DP is fine-tuned via a PPO-based diffusion policy gradient to improve task success rate, while the controller is simultaneously updated to accurately track the planner's evolving command distribution, reducing the distributional mismatch that degrades motion quality. We validate REFINE-DP on a humanoid robot performing loco-manipulation tasks, including door traversal and long-horizon object transport. REFINE-DP achieves an over $90\%$ success rate in simulation, even in out-of-distribution cases not seen in the pre-trained data, and enables smooth autonomous task execution in real-world dynamic environments. Our proposed method substantially outperforms pre-trained DP baselines and demonstrates that RL fine-tuning is key to reliable humanoid loco-manipulation. https://refine-dp.github.io/REFINE-DP/

Time series forecasting has attracted significant attention in the field of AI. Previous works have revealed that the Channel-Independent (CI) strategy improves forecasting performance by modeling each channel individually, but it often suffers from poor generalization and overlooks meaningful inter-channel interactions. Conversely, Channel-Dependent (CD) strategies aggregate all channels, which may introduce irrelevant information and lead to oversmoothing. Despite recent progress, few existing methods offer the flexibility to adaptively balance CI and CD strategies in response to varying channel dependencies. To address this, we propose a generic plugin xCPD, that can adaptively model the channel-patch dependencies from the perspective of graph spectral decomposition. Specifically, xCPD first projects multivariate signals into the frequency domain using a shared graph Fourier basis, and groups patches into low-, mid-, and high-frequency bands based on their spectral energy responses. xCPD then applies a channel-adaptive routing mechanism that dynamically adjusts the degree of inter-channel interaction for each patch, enabling selective activation of frequency-specific experts. This facilitates fine-grained input-aware modeling of smooth trends, local fluctuations, and abrupt transitions. xCPD can be seamlessly integrated on top of existing CI and CD forecasting models, consistently enhancing both accuracy and generalization across benchmarks. The code is available https://github.com/Clearloveyuan/xCPD.

While powered wheelchairs reduce physical fatigue as opposed to manual wheelchairs for individuals with mobility impairment, they demand high cognitive workload due to information processing, decision making and motor coordination. Current autonomous systems lack social awareness in navigation and transparency in decision-making, leading to decreased perceived safety and trust from the user and others in context. This work proposes Socially Aware Autonomous Transparent Transportation (SAATT) Navigation framework for wheelchairs as a potential solution. By implementing a Large Language Model (LLM) informed of user intent and capable of predicting other peoples' intent as a decision-maker for its local controller, it is able to detect and navigate social situations, such as passing pedestrians or a pair conversing. Furthermore, the LLM textually communicates its reasoning at each waypoint for transparency. In this experiment, it is compared against a standard global planner, a representative competing social navigation model, and an Ablation study in three simulated environments varied by social levels in eight metrics categorized under Safety, Social Compliance, Efficiency, and Comfort. Overall, SAATT Nav outperforms in most social situations and equivalently or only slightly worse in the remaining metrics, demonstrating the potential of a socially aware and transparent autonomous navigation system to assist wheelchair users.

We present the first systematic evaluation of mutual exclusivity (ME) -- the bias to map novel words to novel referents -- in text-only language models trained on child-directed speech. We operationalise ME as referential suppression: when a familiar object is relabelled in a two-referent discourse context, ME predicts decreased probability of the labelled noun at a subsequent completion position. Three pilot findings motivate a pre-registered scale-sensitivity experiment: (1) a masked language model (BabyBERTa) is entirely insensitive to multi-sentence referential context; (2) autoregressive models show robust repetition priming -- the opposite of ME -- when familiar nouns are re-labelled; and (3) a novel context-dependence diagnostic reveals that apparent ME-like patterns with nonce tokens are fully explained by embedding similarity, not referential disambiguation. In the confirmatory experiment, we train 45 GPT-2-architecture models (2.9M, 8.9M, and 33.5M parameters; 5, 10, and 20 epochs on AO-CHILDES; 5 seeds each) and evaluate on a pre-registered ME battery. Anti-ME repetition priming is significant in all 9 cells (85-100% of items; all p < 2.4 x 10^-13). Priming attenuates with improved language modelling (Spearman rho = -0.533, p = 0.0002) but never crosses zero across a 3.8x perplexity range. The context-dependence diagnostic replicates in all 9 cells, and dose-response priming increases with repetitions in 8/9 cells (all trend p < 0.002). These findings indicate that distributional learning on child-directed speech produces repetition-based reference tracking rather than lexical exclusivity. We connect this to the grounded cognition literature and argue that referential grounding may be a necessary ingredient for ME -- an empirical claim about required input structure, not a nativist one.

While Large Language Models (LLMs) excel on standardized medical exams, high scores often fail to translate to high-quality responses for real-world medical queries. Current evaluations rely heavily on multiple-choice questions, failing to capture the unstructured, ambiguous, and long-tail complexities inherent in genuine user inquiries. To bridge this gap, we introduce QuarkMedBench, an ecologically valid benchmark tailored for real-world medical LLM assessment. We compiled a massive dataset spanning Clinical Care, Wellness Health, and Professional Inquiry, comprising 20,821 single-turn queries and 3,853 multi-turn sessions. To objectively evaluate open-ended answers, we propose an automated scoring framework that integrates multi-model consensus with evidence-based retrieval to dynamically generate 220,617 fine-grained scoring rubrics (~9.8 per query). During evaluation, hierarchical weighting and safety constraints structurally quantify medical accuracy, key-point coverage, and risk interception, effectively mitigating the high costs and subjectivity of human grading. Experimental results demonstrate that the generated rubrics achieve a 91.8% concordance rate with clinical expert blind audits, establishing highly dependable medical reliability. Crucially, baseline evaluations on this benchmark reveal significant performance disparities among state-of-the-art models when navigating real-world clinical nuances, highlighting the limitations of conventional exam-based metrics. Ultimately, QuarkMedBench establishes a rigorous, reproducible yardstick for measuring LLM performance on complex health issues, while its framework inherently supports dynamic knowledge updates to prevent benchmark obsolescence.

Reliable flood detection is critical for disaster management, yet classical deep learning models often struggle with the high-dimensional, nonlinear complexities inherent in remote sensing data. To mitigate these limitations, we introduced a novel Quantum-Enhanced Vision Transformer (ViT) that synergizes the global context-awareness of transformers with the expressive feature extraction capabilities of quantum computing. Using remote sensing imagery, we developed a hybrid architecture that processes inputs through parallel pathways, a ViT backbone and a quantum branch utilizing a 4-qubit parameterized quantum circuit for localized feature mapping. These distinct representations were fused to optimize binary classification. Results showed that the proposed hybrid model significantly outperformed a classical ViT baseline, increased overall accuracy from 84.48% to 94.47% and the F1-score from 0.841 to 0.944. Notably, the quantum integration substantially improved discriminative power in complex terrains for both class. These findings validate the potential of quantum-classical hybrid models to enhance precision in hydrological monitoring and earth observation applications.

AI systems increasingly assist human decision making by producing preliminary assessments of complex inputs. However, such AI-generated assessments can often be noisy or systematically biased, raising a central question: how should costly human effort be allocated to correct AI outputs where it matters the most for the final decision? We propose a general decision-theoretic framework for human-AI collaboration in which AI assessments are treated as factor-level signals and human judgments as costly information that can be selectively acquired. We consider cases where the optimal selection problem reduces to maximizing a reward associated with each candidate subset of factors, and turn policy design into reward estimation. We develop estimation procedures under both nonparametric and linear models, covering contextual and non-contextual selection rules. In the linear setting, the optimal rule admits a closed-form expression with a clear interpretation in terms of factor importance and residual variance. We apply our framework to AI-assisted peer review. Our approach substantially outperforms LLM-only predictions and achieves performance comparable to full human review while using only 20-30% of the human information. Across different selection rules, we find that simpler rules derived under linear models can significantly reduce computational cost without harming final prediction performance. Our results highlight both the value of human intervention and the efficiency of principled dispatching.

Full-duplex voice agents--systems that listen and speak simultaneously--are rapidly moving from research to production. However, existing evaluations address conversational dynamics and task completion in isolation. We introduce $τ$-voice, a benchmark for evaluating voice agents on grounded tasks with real-world complexity: agents must navigate complex multi-turn conversations, adhere to domain policies, and interact with the environment. The framework extends $τ^2$-bench into a novel voice agent benchmark combining verifiable completion of complex grounded tasks, full-duplex interaction, and realistic audio--enabling direct comparison between voice and text performance. A controllable and realistic voice user simulator provides diverse accents, realistic audio environments, and rich turn-taking dynamics; by decoupling simulation from wall-clock time, the user simulator can use the most capable LLM without real-time constraints. We evaluate task completion (pass@1) and voice interaction quality across 278 tasks: while GPT-5 (reasoning) achieves 85%, voice agents reach only 31--51% under clean conditions and 26--38% under realistic conditions with noise and diverse accents--retaining only 30--45% of text capability; qualitative analysis confirms 79--90% of failures stem from agent behavior, suggesting that observed failures primarily reflect agent behavior under our evaluation setup. $τ$-voice provides a reproducible testbed for measuring progress toward voice agents that are natural, conversational, and reliable.

Although debiased LLMs perform well on known bias patterns, they often fail to generalize to unfamiliar bias prompts, producing toxic outputs. We first validate that such high-bias prompts constitute a \emph{distribution shift} via OOD detection, and show static models degrade under this shift. To adapt on-the-fly, we propose \textbf{CAP-TTA}, a test-time adaptation framework that performs context-aware LoRA updates only when the bias-risk \emph{trigger} exceeds a threshold, using a precomputed diagonal \emph{preconditioner} for fast and stable updates. Across toxic-prompt settings and benchmarks, CAP-TTA reduces bias (confirmed by human evaluation) while achieving much lower update latency than AdamW/SGD; it also mitigates catastrophic forgetting by significantly improving narrative fluency over SOTA debiasing baseline while maintaining comparable debiasing effectiveness.

PET theranostics is transforming precision oncology, yet treatment response varies substantially; many patients receiving 177Lu-PSMA radioligand therapy (RLT) for metastatic castration-resistant prostate cancer (mCRPC) fail to respond, demanding reliable pre-therapy prediction. While LLM-based agents have shown remarkable potential in complex medical diagnosis, their application to PET theranostic outcome prediction remains unexplored, which faces three key challenges: (1) data and knowledge scarcity: RLT was only FDA-approved in 2022, yielding few training cases and insufficient domain knowledge in general LLMs; (2) heterogeneous information integration: robust prediction hinges on structured knowledge extraction from PET/CT, laboratory tests, and free-text clinical documentation; (3) evidence-grounded reasoning: clinical decisions must be anchored in trial evidence rather than LLM hallucinations. In this paper, we present TheraAgent, to our knowledge, the first agentic framework for PET theranostics, with three core innovations: (1) Multi-Expert Feature Extraction with Confidence-Weighted Consensus, where three specialized experts process heterogeneous inputs with uncertainty quantification; (2) Self-Evolving Agentic Memory (SEA-Mem), which learns prognostic patterns from accumulated cases, enabling case-based reasoning from limited data; (3) Evidence-Calibrated Reasoning, integrating a curated theranostics knowledge base to ground predictions in VISION/TheraP trial evidence. Evaluated on 35 real patients and 400 synthetic cases, TheraAgent achieves 75.7% overall accuracy on real patients and 87.0% on synthetic cases, outperforming MDAgents and MedAgent-Pro by over 20%. These results highlight a promising blueprint for trustworthy AI agents in PET theranostics, enabling trial-calibrated, multi-source decision support. Code will be released upon acceptance.

Graphics processing units (GPUs) power many intelligent transportation systems (ITS) and automated driving applications, but remain largely unmonitored for safety and security. This article highlights GPU misuse as a critical blind spot, showing how unmanaged GPU workloads silently degrade real-time performance, demonstrating the need for stronger security measures in ITS.

Most machine learning methods assume fixed probability distributions, limiting their applicability in nonstationary real-world scenarios. While continual learning methods address this issue, current approaches often rely on black-box models or require extensive user intervention for interpretability. We propose SyMPLER (Systems Modeling through Piecewise Linear Evolving Regression), an explainable model for time series forecasting in nonstationary environments based on dynamic piecewise-linear approximations. Unlike other locally linear models, SyMPLER uses generalization bounds from Statistical Learning Theory to automatically determine when to add new local models based on prediction errors, eliminating the need for explicit clustering of the data. Experiments show that SyMPLER can achieve comparable performance to both black-box and existing explainable models while maintaining a human-interpretable structure that reveals insights about the system's behavior. In this sense, our approach conciliates accuracy and interpretability, offering a transparent and adaptive solution for forecasting nonstationary time series.

Accurate extraction of Alzheimer's Disease and Related Dementias (ADRD) phenotypes from electronic health records (EHR) is critical for early-stage detection and disease staging. However, this information is usually embedded in unstructured textual data rather than tabular data, making it difficult to be extracted accurately. We therefore propose LLM-MINE, a Large Language Model-based phenotype mining framework for automatic extraction of ADRD phenotypes from clinical notes. Using two expert-defined phenotype lists, we evaluate the extracted phenotypes by examining their statistical significance across cohorts and their utility for unsupervised disease staging. Chi-square analyses confirm statistically significant phenotype differences across cohorts, with memory impairment being the strongest discriminator. Few-shot prompting with the combined phenotype lists achieves the best clustering performance (ARI=0.290, NMI=0.232), substantially outperforming biomedical NER and dictionary-based baselines. Our results demonstrate that LLM-based phenotype extraction is a promising tool for discovering clinically meaningful ADRD signals from unstructured notes.

Machine learning is now a central part of how modern systems are built and used, powering everything from personalized recommendations to large-scale business analytics. As its role grows, organizations are facing new challenges in managing, deploying, and scaling these models efficiently. One approach that has gained wide adoption is the use of microservice architectures, which break complex machine learning systems into smaller, independent parts that can be built, updated, and scaled on their own. In this paper, we review how major companies such as Netflix, Uber, and Google use microservices to handle key machine learning tasks like training, deployment, and monitoring. We discuss the main challenges involved in designing such systems and explore how microservices fit into large-scale applications, particularly in recommendation systems. We also present some simulation studies showing that microservice-based designs can reduce latency and improve scalability, leading to faster, more efficient, and more responsive machine learning applications in real-world and large-scale systems.

Global cryptocurrencies are unbacked and have high transaction cost incurred by global consensus. In contrast, grassroots cryptocurrencies are backed by the goods and services of their issuers -- any person, natural or legal -- and have no transaction cost beyond operating a smartphone. Liquidity in grassroots cryptocurrencies arises from mutual credit via coin exchange among issuers. However, as grassroots coins are redeemable 1-for-1 against any other grassroots coin, the credit-forming exchange must also be 1-for-1, lest prompt redemption after exchange would leave the parties with undue profit or loss. Thus, grassroots coins are incongruent with liquidity through interest-bearing credit. Here we introduce grassroots bonds, which extend grassroots coins with a maturity date, reframing grassroots coins -- cash -- as mature grassroots bonds. Bond redemption generalises coin redemption, allowing the lending of liquid coins in exchange for interest-bearing future-maturity bonds. We show that digital social contracts -- voluntary agreements among persons, specified, fulfilled, and enforced digitally -- can express the full gamut of financial instruments as the voluntary swap of grassroots bonds, including credit lines, loans, sale of debt, forward contracts, options, and escrow-based instruments, and that classical liquidity ratios are applicable just as well to grassroots bonds. The formal specification presented here was used by AI to derive a working implementation of grassroots bonds in GLP, a concurrent logic programming language implemented in Dart for smartphone deployment. The implementation is illustrated by a running multiagent village market scenario, also implemented in GLP by AI.

No-Reference Image Quality Assessment (NR-IQA) aims to estimate perceptual quality without access to a reference image of pristine quality. Learning an NR-IQA model faces a fundamental bottleneck: its need for a large number of costly human perceptual labels. We propose SHAMISA, a non-contrastive self-supervised framework that learns from unlabeled distorted images by leveraging explicitly structured relational supervision. Unlike prior methods that impose rigid, binary similarity constraints, SHAMISA introduces implicit structural associations, defined as soft, controllable relations that are both distortion-aware and content-sensitive, inferred from synthetic metadata and intrinsic feature structure. A key innovation is our compositional distortion engine, which generates an uncountable family of degradations from continuous parameter spaces, grouped so that only one distortion factor varies at a time. This enables fine-grained control over representational similarity during training: images with shared distortion patterns are pulled together in the embedding space, while severity variations produce structured, predictable shifts. We integrate these insights via dual-source relation graphs that encode both known degradation profiles and emergent structural affinities to guide the learning process throughout training. A convolutional encoder is trained under this supervision and then frozen for inference, with quality prediction performed by a linear regressor on its features. Extensive experiments on synthetic, authentic, and cross-dataset NR-IQA benchmarks demonstrate that SHAMISA achieves strong overall performance with improved cross-dataset generalization and robustness, all without human quality annotations or contrastive losses.

Despite advances in assistive technologies, Blind and Low-Vision (BLV) individuals continue to face challenges in understanding their surroundings. Delivering concise, useful, and timely scene descriptions for ambient perception remains a long-standing accessibility problem. To address this, we introduce Audo-Sight, an AI-driven assistive system across Edge-Cloud that enables BLV individuals to perceive their surroundings through voice-based conversational interaction. Audo-Sight employs a set of expert and generic AI agents, each supported by dedicated processing pipelines distributed across edge and cloud. It analyzes user queries by considering urgency and contextual information to infer the user intent and dynamically route each query, along with a scene frame, to the most suitable pipeline. In cases where users require fast responses, the system simultaneously leverages edge and cloud processing pipelines. The edge generates an initial response quickly, while the cloud provides more detailed and accurate information. To overcome the challenge of seamlessly combining these outputs, we introduce the Response Fusion Engine, which fuses the fast edge response with the more accurate cloud output, ensuring timely and high-accuracy response for the BLV users. Systematic evaluation shows that Audo-Sight delivers speech output around 80% faster for urgent tasks and generates complete responses approximately 50% faster across all tasks compared to a commercial cloud-based solution -- highlighting the effectiveness of our system across edge-cloud. Human evaluation of Audo-Sight shows that it is the preferred choice over GPT-5 for 62% of BLV participants with another 23% stating both perform comparably.

In this work, we propose an unsupervised domain adaptation (UDA) framework for 3D volumetric lesion detection that adapts a detector trained on labeled FDG PET/CT to unlabeled PSMA PET/CT. Beyond covariate shift, cross tracer adaptation also exhibits label shift in both lesion size composition and the number of lesions per subject. We introduce self-training with two mechanisms that explicitly model and compensate for this label shift. First, we adaptively adjust the detection anchor shapes by re-estimating target domain box scales from selected pseudo labels and updating anchors with an exponential moving average. This increases positive anchor coverage for small PSMA lesions and stabilizes box regression. Second, instead of a fixed confidence threshold for pseudo-label selection, we allocate size bin-wise quotas according to the estimated target domain histogram over lesion volumes. The self-training alternates between supervised learning with prior-guided pseudo labeling on PSMA and supervised learning on labeled FDG. On AutoPET 2024, adapting from 501 labeled FDG studies to 369 $^{18}$F-PSMA studies, the proposed method improves both AP and FROC over the source-only baseline and conventional self-training without label-shift mitigation, indicating that modeling target lesion prevalence and size composition is an effective path to robust cross-tracer detection.

Advanced nodes decouple contacted poly pitch (CPP) and lower-metal pitch to improve routability. We present CPCell, an efficient standard-cell layout generation framework, to support arbitrary gear ratio (GR) and offset parameters through a fine-grained layered grid graph and constraint-programming-based placement-routing co-optimization. Layout quality is improved via Middle-of-Line routing, M0 pin enablement, pin accessibility constraints and a weighted multi-objective formulation that jointly optimizes cell layouts. To scale to netlists with up to 48 transistors, we incorporate acceleration techniques including transistor clustering, identical transistor partitioning, routing lower bound tightening and early termination strategies. Comprehensive cell-level and block-level studies are conducted to evaluate GR and offset choices, quantify the benefits of the proposed objectives and assess their impact on power, performance, area and IR-drop outcomes.

The success of vision transformers-especially for generative modeling-is limited by the quadratic cost and weak spatial inductive bias of self-attention. We propose PDE-SSM, a spatial state-space block that replaces attention with a learnable convection-diffusion-reaction partial differential equation. This operator encodes a strong spatial prior by modeling information flow via physically grounded dynamics rather than all-to-all token interactions. Solving the PDE in the Fourier domain yields global coupling with near-linear complexity of $O(N \log N)$, delivering a principled and scalable alternative to attention. We integrate PDE-SSM into a flow-matching generative model to obtain the PDE-based Diffusion Transformer PDE-SSM-DiT. Empirically, PDE-SSM-DiT matches or exceeds the performance of state-of-the-art Diffusion Transformers while substantially reducing compute. Our results show that, analogous to 1D settings where SSMs supplant attention, multi-dimensional PDE operators provide an efficient, inductive-bias-rich foundation for next-generation vision models.

The recent escalation of the Iran Israel USA conflict in 2026 has triggered widespread global discussions across social media platforms. As people increasingly use these platforms for expressing opinions, analyzing public sentiment from these discussions can provide valuable insights into global public perception. This study aims to analyze global public sentiment regarding the Iran Israel USA conflict by mining user-generated comments from YouTube news channels. The work contributes to public opinion analysis by introducing a privacy preserving framework that combines topic wise sentiment analysis with modern deep learning techniques and Federated Learning. To achieve this, approximately 19,000 YouTube comments were collected from major international news channels and preprocessed to remove noise and normalize text. Sentiment labels were initially generated using the VADER sentiment analyzer and later validated through manual inspection to improve reliability. Latent Dirichlet Allocation (LDA) was applied to identify key discussion topics related to the conflict. Several transformer-based models, including BERT, RoBERTa, XLNet, DistilBERT, ModernBERT, and ELECTRA, were fine tuned for sentiment classification. The best-performing model was further integrated into a federated learning environment to enable distributed training by preserving user data privacy. Additionally, Explainable Artificial Intelligence (XAI) techniques using SHAP were applied to interpret model predictions and identify influential words affecting sentiment classification. Experimental results demonstrate that transformer models perform effectively, and among them, ELECTRA achieved the best performance with 91.32% accuracy. The federated learning also maintained strong performance while preserving privacy, achieving 89.59% accuracy in a two client configuration.

Bibliographic reference extraction and parsing are foundational for citation indexing, linking, and downstream scholarly knowledge-graph construction. However, most established evaluations focus on clean, English, end-of-document bibliographies, and therefore underrepresent the Social Sciences and Humanities (SSH), where citations are frequently multilingual, embedded in footnotes, abbreviated, and shaped by heterogeneous historical conventions. We present a unified benchmark that targets these SSH-realistic conditions across three complementary datasets: CEX (English journal articles spanning multiple disciplines), EXCITE (German/English documents with end-section, footnote-only, and mixed regimes), and LinkedBooks (humanities references with strong stylistic variation and multilinguality). We evaluate three tasks of increasing difficulty -- reference extraction, reference parsing, and end-to-end document parsing -- under a schema-constrained setup that enables direct comparison between a strong supervised pipeline baseline (GROBID) and contemporary LLMs (DeepSeek-V3.1, Mistral-Small-3.2-24B, Gemma-3-27B-it, and Qwen3-VL (4B-32B variants)). Across datasets, extraction largely saturates beyond a moderate capability threshold, while parsing and end-to-end parsing remain the primary bottlenecks due to structured-output brittleness under noisy layouts. We further show that lightweight LoRA adaptation yields consistent gains -- especially on SSH-heavy benchmarks -- and that segmentation/pipelining can substantially improve robustness. Finally, we argue for hybrid deployment via routing: leveraging GROBID for well-structured, in-distribution PDFs while escalating multilingual and footnote-heavy documents to task-adapted LLMs.

Foundation models succeed when they learn in the native structure of a modality, whether morphology-respecting tokens in language or pixels in vision. Wireless packet traces deserve the same treatment: meaning emerges from layered headers, typed fields, timing gaps, and cross-packet state machines, not flat strings. We present Plume (Protocol Language Understanding Model for Exchanges), a compact 140M-parameter foundation model for 802.11 traces that learns from structured PDML dissections. A protocol-aware tokenizer splits along the dissector field tree, emits gap tokens for timing, and normalizes identifiers, yielding 6.2x shorter sequences than BPE with higher per token information density. Trained on a curated corpus, Plume achieves 74-97% next-packet token accuracy across five real-world failure categories and AUROC >= 0.99 for zero-shot anomaly detection. On the same prediction task, frontier LLMs (Claude Opus 4.6, GPT-5.4) score comparably despite receiving identical protocol context, yet Plume does so with > 600x fewer parameters, fitting on a single GPU at effectively zero marginal cost vs. cloud API pricing, enabling on-prem, privacy-preserving root cause analysis.

Large language models (LLMs) and small language models (SLMs) operate under strict context window and key-value (KV) cache constraints, fundamentally limiting their ability to reason coherently over long interaction horizons. Existing approaches -- extended context windows, retrieval-augmented generation, summarization, or static documentation -- treat memory as static storage and fail to preserve decision-relevant state under long-running, multi-session tasks. We introduce StatePlane, a model-agnostic cognitive state plane that governs the formation, evolution, retrieval, and decay of episodic, semantic, and procedural state for AI systems operating under bounded context. Grounded in cognitive psychology and systems design, StatePlane formalizes episodic segmentation, selective encoding via information-theoretic constraints, goal-conditioned retrieval with intent routing, reconstructive state synthesis, and adaptive forgetting. We present a formal state model, KV-aware algorithms, security and governance mechanisms including write-path anti-poisoning, enterprise integration pathways, and an evaluation framework with six domain-specific benchmarks. StatePlane demonstrates that long-horizon intelligence can be achieved without expanding context windows or retraining models.

Code transformation is a foundational capability in the software development process, where its effectiveness relies on constructing a high-quality code representation to characterize the input code semantics and guide the transformation. Existing approaches treat code transformation as an end-to-end learning task, leaving the construction of the representation needed for semantic reasoning implicit in model weights or relying on rigid compiler-level abstractions. We present SemRep, a framework that improves code transformation through generative code representation learning. Our key insight is to employ the semantics-preserving transformations as the intermediate representation, which serves as both a generative mid-training task and the guidance for subsequent instruction-specific code transformations. Across general code editing and optimization tasks (e.g., GPU kernel optimization), SemRep outperforms the extensively finetuned baselines with strictly the same training budget by 6.9% in correctness, 1.1x in performance, 13.9% in generalization, and 6.7% in robustness. With the improved exploration of diverse code transformations, SemRep is particularly amenable to evolutionary search. Combined with an evolutionary coding agent, SemRep finds optimizations that 685B larger-weight baselines fail to discover while achieving the same performance with 25% less inference compute.

Static information presentation in VR cultural heritage often causes cognitive overload or under-stimulation. We introduce a closed-loop adaptive interface that tailors content depth to real-time visitor behavior through implicit multimodal sensing. Our approach continuously monitors gaze dwell, head kinematics, and locomotion to infer engagement via a transparent rule-based classifier, which drives a Large Language Model to dynamically modulate explanation complexity without interrupting exploration. We implemented a proof-of-concept in the Berat Ethnographic Museum and conducted a preliminary evaluation (N=16) comparing adaptive versus static content. Results indicate that adaptive participants demonstrated 2-3x increases in reading engagement and exploration time while maintaining high usability (SUS = 84.3). Technical validation confirmed sub-millisecond engagement inference latency on consumer VR hardware. These preliminary findings warrant larger-scale investigation and raise questions about engagement validation, AI transparency, and generative models in heritage contexts. We present this work-in-progress to spark discussion about implicit AI-driven adaptation in immersive cultural experiences.

Large language models (LLMs) are increasingly used to assess moral or ethical statements, yet their judgments may reflect social and linguistic biases. This work presents a controlled, sentence-level study of how grammatical person, number, and gender markers influence LLM moral classifications of fairness. Starting from 550 balanced base sentences from the ETHICS dataset, we generated 26 counterfactual variants per item, systematically varying pronouns and demographic markers to yield 14,850 semantically equivalent sentences. We evaluated six model families (Grok, GPT, LLaMA, Gemma, DeepSeek, and Mistral), and measured fairness judgments and inter-group disparities using Statistical Parity Difference (SPD). Results show statistically significant biases: sentences written in the singular form and third person are more often judged as "fair'', while those in the second person are penalized. Gender markers produce the strongest effects, with non-binary subjects consistently favored and male subjects disfavored. We conjecture that these patterns reflect distributional and alignment biases learned during training, emphasizing the need for targeted fairness interventions in moral LLM applications.

The emergence of Vision-Language Models (VLMs) has introduced new paradigms for global image geo-localization through retrieval-augmented generation (RAG) and reasoning-driven inference. However, RAG methods are constrained by retrieval database quality, while reasoning-driven approaches fail to internalize image locatability, relying on inefficient, fixed-depth reasoning paths that increase hallucinations and degrade accuracy. To overcome these limitations, we introduce an Optimized Locatability Score that quantifies an image's suitability for deep reasoning in geo-localization. Using this metric, we curate Geo-ADAPT-51K, a locatability-stratified reasoning dataset enriched with augmented reasoning trajectories for complex visual scenes. Building on this foundation, we propose a two-stage Group Relative Policy Optimization (GRPO) curriculum with customized reward functions that regulate adaptive reasoning depth, visual grounding, and hierarchical geographical accuracy. Our framework, Geo-ADAPT, learns an adaptive reasoning policy, achieves state-of-the-art performance across multiple geo-localization benchmarks, and substantially reduces hallucinations by reasoning both adaptively and efficiently.

Chemical language models (CLMs) have emerged as promising competitors to popular classical machine learning models for molecular property prediction (MPP) tasks. However, an increasing number of studies have reported inconsistent and contradictory results for the performance of CLMs across various MPP benchmark tasks. In this study, we conduct and analyze hundreds of meticulously controlled experiments to systematically investigate the effects of various factors, such as dataset size, model size, and standardization, on the pre-training and fine-tuning performance of CLMs for MPP. In the absence of well-established scaling laws for encoder-only masked language models, our aim is to provide comprehensive numerical evidence and a deeper understanding of the underlying mechanisms affecting the performance of CLMs for MPP tasks, some of which appear to be entirely overlooked in the literature.

Twitter (now X) has become an important source of social media data for situational awareness during crises. Crisis informatics research has widely used tweets from Twitter to develop and evaluate artificial intelligence (AI) systems for various crisis-relevant tasks, such as extracting locations and estimating damage levels from tweets to support damage assessment. However, recent changes in Twitter's data access policies have made it increasingly difficult to curate real-world tweet datasets related to crises. Moreover, existing curated tweet datasets are limited to past crisis events in specific contexts and are costly to annotate at scale. These limitations constrain the development and evaluation of AI systems used in crisis informatics. To address these limitations, we introduce an agentic workflow for generating crisis-related synthetic tweet datasets. The workflow iteratively generates synthetic tweets conditioned on prespecified target characteristics, evaluates them using predefined compliance checks, and incorporates structured feedback to refine them in subsequent iterations. As a case study, we apply the workflow to generate synthetic tweet datasets relevant to post-earthquake damage assessment. We show that the workflow can generate synthetic tweets that capture their target labels for location and damage level. We further demonstrate that the resulting synthetic tweet datasets can be used to evaluate AI systems on damage assessment tasks like geolocalization and damage level prediction. Our results indicate that the workflow offers a flexible and scalable alternative to real-world tweet data curation, enabling the systematic generation of synthetic social media data across diverse crisis events, societal contexts, and crisis informatics applications.

Fraud-related financial losses continue to rise, while regulatory, privacy, and data-sovereignty constraints increasingly limit the feasibility of centralized fraud detection systems. Federated Learning (FL) has emerged as a promising paradigm for enabling collaborative model training across institutions without sharing raw transaction data. Yet, its practical effectiveness under realistic, non-IID financial data distributions remains insufficiently validated. In this work, we present a multi-institution, industry-oriented proof-of-concept study evaluating federated anomaly detection for payment transactions using the NVIDIA FLARE framework. We simulate a realistic federation of heterogeneous financial institutions, each observing distinct fraud typologies and operating under strict data isolation. Using a deep neural network trained via federated averaging (FedAvg), we demonstrate that federated models achieve a mean F1-score of 0.903 - substantially outperforming locally trained models (0.643) and closely approaching centralized training performance (0.925), while preserving full data sovereignty. We further analyze convergence behavior, showing that strong performance is achieved within 10 federated communication rounds, highlighting the operational viability of FL in latency- and cost-sensitive financial environments. To support deployment in regulated settings, we evaluate model interpretability using Shapley-based feature attribution and confirm that federated models rely on semantically coherent, domain-relevant decision signals. Finally, we incorporate sample-level differential privacy via DP-SGD and demonstrate favorable privacy-utility trade-offs...

Standard sequential inference architectures are compromised by a normalizability crisis when confronted with extreme, structured outliers. By operating on unbounded parameter spaces, state-of-the-art estimators lack the intrinsic geometry required to appropriately sever anomalies, resulting in unbounded covariance inflation and mean divergence. This paper resolves this structural failure by analyzing the abstraction sequence of inference at the meta-prior level (S_2). We demonstrate that extremizing the action over an infinite-dimensional space requires a non-parametric field anchored by a pre-prior, as a uniform volume element mathematically does not exist. By utilizing strictly invariant Delta (or ν) Information Separations on the statistical manifold, we physically truncate the infinite tails of the spatial distribution. When evaluated as a Radon-Nikodym derivative against the base measure, the active parameter space compresses into a strictly finite, normalizable probability droplet. Empirical benchmarks across three domains--LiDAR maneuvering target tracking, high-frequency cryptocurrency order flow, and quantum state tomography--demonstrate that this bounded information geometry analytically truncates outliers, ensuring robust estimation without relying on infinite-tailed distributional assumptions.

Routing a query through an appropriate LLM is challenging, particularly when user preferences are expressed in natural language and model attributes are only partially observable. We propose a constraint-based interpretation of language-conditioned LLM routing, formulating it as a weighted MaxSAT/MaxSMT problem in which natural language feedback induces hard and soft constraints over model attributes. Under this view, routing corresponds to selecting models that approximately maximize satisfaction of feedback-conditioned clauses. Empirical analysis on a 25-model benchmark shows that language feedback produces near-feasible recommendation sets, while no-feedback scenarios reveal systematic priors. Our results suggest that LLM routing can be understood as structured constraint optimization under language-conditioned preferences.

Mixture-of-Experts (MoE) architectures have become essential for scaling large language models, driving the development of specialized device-initiated communication libraries such as DeepEP, Hybrid-EP, and others. These libraries demonstrate the performance benefits of GPU-initiated RDMA for MoE dispatch and combine operations. This paper presents NCCL EP (Expert Parallelism), a ground-up MoE communication library built entirely on NCCL's Device API. NCCL EP provides unified ncclEpDispatch and ncclEpCombine primitives with both C and Python interfaces, supporting Low-Latency (LL) mode for inference decoding and High-Throughput (HT) mode for training and inference prefill. LL targets small batch sizes (1-128 tokens) using direct all-to-all RDMA+NVLink mesh connectivity with double-buffered communication for overlapping dispatch and combine phases. HT targets large batches (4096+ tokens) using hierarchical communication that aggregates tokens within NVLink domains before inter-node RDMA transmission. Both modes leverage Device API for both intra- and inter-node communications, taking advantage of its topology awareness and optimized GPU-initiated implementation. We evaluate NCCL EP on an H100-based cluster across multi-node configurations, demonstrating competitive LL kernel performance and presenting end-to-end results with vLLM integration. By building MoE communication natively within NCCL, NCCL EP provides a supported path for expert parallelism on current and emerging NVIDIA platforms.

We introduce Orla, a library for constructing and running LLM-based agentic systems. Modern agentic applications consist of workflows that combine multiple LLM inference steps, tool calls, and heterogeneous infrastructure. Today, developers typically build these systems by manually composing orchestration code with LLM serving engines and tool execution logic. Orla provides a general abstraction that separates request execution from workflow-level policy. It acts as a serving layer above existing LLM inference engines: developers define workflows composed of stages, while Orla manages how those stages are mapped, executed, and coordinated across models and backends. It provides agent-level control through three mechanisms: a stage mapper, which assigns each stage to an appropriate model and backend; a workflow orchestrator, which schedules stages and manages their resources and context; and a memory manager, which manages inference state such as the KV cache across workflow boundaries. We demonstrate Orla with a customer support workflow that exercises many of its capabilities. We evaluate Orla on two datasets, showing that stage mapping improves latency and cost compared to a single-model vLLM baseline, while workflow-level cache management reduces time-to-first-token.

We prove The Equivalence Theorem: structurally complete knowledge representation requires exactly four mutually entailing capabilities -- n-ary relationships with attributes, temporal validity, uncertainty quantification, and causal relationships between relationships -- collectively equivalent to treating relationships as first-class objects. Any system implementing one capability necessarily requires all four; any system missing one cannot achieve structural completeness. This result is constructive: we exhibit an Attributed Temporal Causal Hypergraph (ATCH) framework satisfying all four conditions simultaneously. The theorem yields a strict expressiveness hierarchy -- SQL < LPG < TypeDB < ATCH -- with witness queries that are structurally inexpressible at each lower level. We establish computational complexity bounds showing NP-completeness for general queries but polynomial-time tractability for practical query classes (acyclic patterns, bounded-depth causal chains, windowed temporal queries). As direct corollaries, we derive solutions to classical AI problems: the Frame Problem (persistence by default from temporal validity), conflict resolution (contradictions as unresolved metadata with hidden variable discovery), and common sense reasoning (defaults with causal inhibitors). A prototype PostgreSQL extension in C validates practical feasibility within the established complexity bounds.

Developing predictive models that perform reliably across diverse patient populations and heterogeneous environments is a core aim of medical research. However, generalization is only possible if the learned model is robust to statistical differences between data used for training and data seen at the time and place of deployment. Domain generalization methods provide strategies to address data shifts, but each method comes with its own set of assumptions and trade-offs. To apply these methods in healthcare, we must understand how domain shifts arise, what assumptions we prefer to make, and what our design constraints are. This article proposes a causal framework for the design of predictive models to improve generalization. Causality provides a powerful language to characterize and understand diverse domain shifts, regardless of data modality. This allows us to pinpoint why models fail to generalize, leading to more principled strategies to prepare for and adapt to shifts. We recommend general mitigation strategies, discussing trade-offs and highlighting existing work. Our causality-based perspective offers a critical foundation for developing robust, interpretable, and clinically relevant AI solutions in healthcare, paving the way for reliable real-world deployment.

Large language models are shifting from passive information providers to active agents intended for complex workflows. However, their deployment as reliable AI workers in enterprise is stalled by benchmarks that fail to capture the intricacies of professional environments, specifically, the need for long-horizon planning amidst persistent state changes and strict access protocols. In this work, we introduce EnterpriseOps-Gym, a benchmark designed to evaluate agentic planning in realistic enterprise settings. Specifically, EnterpriseOps-Gym features a containerized sandbox with 164 database tables and 512 functional tools to mimic real-world search friction. Within this environment, agents are evaluated on 1,150 expert-curated tasks across eight mission-critical verticals (including Customer Service, HR, and IT). Our evaluation of 14 frontier models reveals critical limitations in state-of-the-art models: the top-performing Claude Opus 4.5 achieves only 37.4% success. Further analysis shows that providing oracle human plans improves performance by 14-35 percentage points, pinpointing strategic reasoning as the primary bottleneck. Additionally, agents frequently fail to refuse infeasible tasks (best model achieves 53.9%), leading to unintended and potentially harmful side effects. Our findings underscore that current agents are not yet ready for autonomous enterprise deployment. More broadly, EnterpriseOps-Gym provides a concrete testbed to advance the robustness of agentic planning in professional workflows.

Cardiovascular diseases are the leading cause of death. Cardiac phenotypes (CPs), e.g., ejection fraction, are the gold standard for assessing cardiac health, but they are derived from cine cardiac magnetic resonance imaging (CMR), which is costly and requires high spatio-temporal resolution. Every magnetic resonance (MR) examination begins with rapid and coarse localizers for scan planning, which are discarded thereafter. Despite non-diagnostic image quality and lack of temporal information, localizers can provide valuable structural information rapidly. In addition to imaging, patient-level information, including demographics and lifestyle, influence the cardiac health assessment. Electrocardiograms (ECGs) are inexpensive, routinely ordered in clinical practice, and capture the temporal activity of the heart. Here, we introduce C-TRIP (Cardiac Tri-modal Representations for Imaging Phenotypes), a multi-modal framework that aligns localizer MRI, ECG signals, and tabular metadata to learn a robust latent space and predict CPs using localizer images as an opportunistic alternative to CMR. By combining these three modalities, we leverage cheap spatial and temporal information from localizers, and ECG, respectively while benefiting from patient-specific context provided by tabular data. Our pipeline consists of three stages. First, encoders are trained independently to learn uni-modal representations. The second stage fuses the pre-trained encoders to unify the latent space. The final stage uses the enriched representation space for CP prediction, with inference performed exclusively on localizer MRI. Proposed C-TRIP yields accurate functional CPs, and high correlations for structural CPs. Since localizers are inherently rapid and low-cost, our C-TRIP framework could enable better accessibility for CP estimation.

In precipitation nowcasting, most extrapolation-based methods rely on two-dimensional radar composites to estimate the horizontal motion of precipitation systems. However, in some cases, precipitation systems can exhibit varying motion at different heights. We propose a physics-informed convolutional neural network that estimates independent horizontal motion fields for multiple altitude layers directly from volumetric radar reflectivity data and investigate the practical benefits of altitude-wise motion field estimation for precipitation nowcasting. The model is trained end-to-end on volumetric observations from the Slovak radar network and its extrapolation nowcasting performance is evaluated. We compare the proposed model against an architecturally identical baseline operating on vertically pooled two-dimensional radar composites. Our results show that, although the model successfully learns altitude-wise motion fields, the estimated displacement is highly correlated across vertical levels for the vast majority of precipitation events. Consequently, the volumetric approach does not yield systematic improvements in nowcasting accuracy. While categorical metrics indicate increased precipitation detection at longer lead times, this gain is largely attributable to non-physical artifacts and is accompanied by a growing positive bias. A comprehensive inter-altitude motion field correlation analysis further confirms that events exhibiting meaningful vertical variability in horizontal motion are rare in the studied region. We conclude that, for the Slovak radar dataset, the additional complexity of three-dimensional motion field estimation is not justified by questionable gains in predictive skill. Nonetheless, the proposed framework remains applicable in climates where precipitation systems exhibit stronger vertical variability in horizontal motion.

State-space models (SSMs) are effective architectures for sequential modeling, but a rigorous theoretical understanding of their training dynamics is still lacking. In this work, we formulate the training of SSMs as an ensemble optimal control problem, where a shared control law governs a population of input-dependent dynamical systems. We derive Pontryagin's maximum principle (PMP) for this ensemble control formulation, providing necessary conditions for optimality. Motivated by these conditions, we introduce an algorithm based on the method of successive approximations. We prove convergence of this iterative scheme along a subsequence and establish sufficient conditions for global optimality. The resulting framework provides a control-theoretic perspective on SSM training.

Deep learning has achieved recognition for its impact within natural sciences, however scientists are inhibited by the prohibitive technical cost and computational complexity of training project specific models from scratch. Following software engineering community guidance, natural scientists are reusing pre-trained deep learning models (PTMs) to amortize these costs. While prior works recommend PTM reuse patterns, to our knowledge, little work has been done to empirically evaluate their usage and impact within the natural sciences. We present the first empirical study of PTM reuse patterns in the natural sciences, quantifying the utilization and impact of conceptual, adaptation, and deployment reuse within the scientific process. Leveraging an automated large language model driven pipeline, we analyze 17,511 peer reviewed, open access papers to identify PTM reuse by scientific field, associated reuse patterns, and the impact of PTM integration into the scientific process from January 1st, 2000 to December 10th, 2025. Our results show that "Biochemistry, Genetics and Molecular Biology" has outpaced other natural scientific fields in PTM reuse, "adaptation" reuse is the most prevalent PTM reuse pattern identified across all natural science fields, and the "Test" stage of the scientific process has been most impacted by PTM integration. This aligns with the growing interest of leveraging computational methods to conduct high throughput, data driven scientific research. Our work characterizes and identifies current PTM reuse practices within the natural sciences, evaluates their impact on the scientific process, and establishes a foundation for future work into the implementation and broader scientific implications of PTM reuse.

This paper presents a Probabilistic State Algebra as an extension of deterministic propositional logic, providing a computational framework for constructing Markov Random Fields (MRFs) through pure linear algebra. By mapping logical states to real-valued coordinates interpreted as energy potentials, we define an energy-based model where global probability distributions emerge from coordinate-wise Hadamard products. This approach bypasses the traditional reliance on graph-traversal algorithms and compiled circuits, utilising $t$-objects and wildcards to embed logical reduction natively within matrix operations. We demonstrate that this algebra constructs formal Gibbs distributions, offering a rigorous mathematical link between symbolic constraints and statistical inference. A central application of this framework is the development of Probabilistic Rule Models (PRMs), which are uniquely capable of incorporating both probabilistic associations and deterministic logical constraints simultaneously. These models are designed to be inherently interpretable, supporting a human-in-the-loop approach to decisioning in high-stakes environments such as healthcare and finance. By representing decision logic as a modular summation of rules within a vector space, the framework ensures that complex probabilistic systems remain auditable and maintainable without compromising the rigour of the underlying configuration space.

The rapid proliferation of the Internet of Things has intensified demand for robust privacy-preserving machine learning mechanisms to safeguard sensitive data generated by large-scale, heterogeneous, and resource-constrained devices. Unlike centralized environments, IoT ecosystems are inherently decentralized, bandwidth-limited, and latency-sensitive, exposing privacy risks across sensing, communication, and distributed training pipelines. These characteristics render conventional anonymization and centralized protection strategies insufficient for practical deployments. This survey presents a comprehensive IoT-centric, cross-paradigm analysis of privacy-preserving machine learning. We introduce a structured taxonomy spanning perturbation-based mechanisms such as differential privacy, distributed paradigms such as federated learning, cryptographic approaches including homomorphic encryption and secure multiparty computation, and generative synthesis techniques based on generative adversarial networks. For each paradigm, we examine formal privacy guarantees, computational and communication complexity, scalability under heterogeneous device participation, and resilience against threats including membership inference, model inversion, gradient leakage, and adversarial manipulation. We further analyze deployment constraints in wireless IoT environments, highlighting trade-offs between privacy, communication overhead, model convergence, and system efficiency within next-generation mobile architectures. We also consolidate evaluation methodologies, summarize representative datasets and open-source frameworks, and identify open challenges including hybrid privacy integration, energy-aware learning, privacy-preserving large language models, and quantum-resilient machine learning.

Imbalanced datasets pose a difficulty in fraud detection, as classifiers are often biased toward the majority class and perform poorly on rare fraudulent transactions. Synthetic data generation is therefore commonly used to mitigate this problem. In this work, we propose the Clustered Embedding Diffusion-Transformer (EmDT), a diffusion model designed to generate fraudulent samples. Our key innovation is to leverage UMAP clustering to identify distinct fraudulent patterns, and train a Transformer denoising network with sinusoidal positional embeddings to capture feature relationships throughout the diffusion process. Once the synthetic data has been generated, we employ a standard decision-tree-based classifier (e.g., XGBoost) for classification, as this type of model remains better suited to tabular datasets. Experiments on a credit card fraud detection dataset demonstrate that EmDT significantly improves downstream classification performance compared to existing oversampling and generative methods, while maintaining comparable privacy protection and preserving feature correlations present in the original data.

Weather forecasting offers an ideal testbed for artificial intelligence (AI) to learn complex, multi-scale physical systems. Traditional numerical weather prediction remains computationally costly for frequent regional updates, as high-resolution nests require intensive boundary coupling. We introduce Multi-Resolution Graph Neural Forecasting (MR-GNF), a lightweight, physics-aware model that performs short-term regional forecasts directly on an ellipsoidal, multi-scale graph of the Earth. The framework couples a 0.25° region of interest with a 0.5° context belt and 1.0° outer domain, enabling continuous cross-scale message passing without explicit nested boundaries. Its axial graph-attention network alternates vertical self-attention across pressure levels with horizontal graph attention across surface nodes, capturing implicit 3-D structure in just 1.6 M parameters. Trained on 40 years of ERA5 reanalysis (1980-2024), MR-GNF delivers stable +6 h to +24 h forecasts for near-surface temperature, wind, and precipitation over the UK-Ireland sector. Despite a total compute cost below 80 GPU-hours on a single RTX 6000 Ada, the model matches or exceeds heavier regional AI systems while preserving physical consistency across scales. These results demonstrate that graph-based neural operators can achieve trustworthy, high-resolution weather prediction at a fraction of NWP cost, opening a practical path toward AI-driven early-warning and renewable-energy forecasting systems. Project page and code: https://github.com/AndriiShchur/MR-GNF

Purpose: Higher education institutions face increasing pressure to audit course designs for generative AI (GenAI) integration. This paper presents an end-to-end method for using large language models (LLMs) to scan course information sheets at scale, identify where assessments may be vulnerable to student use of GenAI tools, validate system performance through iterative refinement, and operationalise results through direct stakeholder communication and effort. Method: We developed a four-phase pipeline: (0) manual pilot sampling, (1) iterative prompt engineering with multi-model comparison, (2) full production scan of 4,684 Bachelor and Master course information sheets (Academic Year 2024-2025) from the Vrije Universiteit Brussel (VUB) with automated report generation and email distribution to teaching teams (91.4% address-matched) using a three-tier risk taxonomy (Clear risk, Potential risk, Low risk), and (3) longitudinal re-scan of 4,675 sheets after the next catalogue release. Results: Five iterations of prompt refinement achieved 87% agreement with expert labels. GPT-4o was selected for production based on superior handling of ambiguous cases involving internships and practical components. The Year 1 scan classified 60.3% of courses as Clear risk, 15.2% as Potential risk, and 24.5% as Low risk. Year 2 comparison revealed substantial shifts in risk distributions, with improvements most pronounced in practice-oriented programmes. Implications: The method enables institutions to rapidly transform heterogeneous catalogue data into structured and actionable intelligence. The approach is transferable to other audit domains (sustainability, accessibility, pedagogical alignment) and provides a template for responsible LLM deployment in higher education governance.

Wireless transmission of high-dimensional 3D point clouds (PCs) is increasingly required in industrial collaborative robotics systems. Conventional compression methods prioritize geometric fidelity, although many practical applications ultimately depend on reliable task-level inference rather than exact coordinate reconstruction. In this paper, we propose an end-to-end semantic communication framework for wireless 3D PC transmission and conduct a systematic study of the relationship between geometric reconstruction fidelity and semantic robustness under channel impairments. The proposed architecture jointly supports geometric recovery and object classification from a shared transmitted representation, enabling direct comparison between coordinate-level and task-level sensitivity to noise. Experimental evaluation on a real industrial dataset reveals a pronounced asymmetry: semantic inference remains stable across a broad signal-to-noise ratio (SNR) range even when geometric reconstruction quality degrades significantly. These results demonstrate that reliable task execution does not require high-fidelity geometric recovery and provide design insights for task-oriented wireless perception systems in bandwidth- and power-constrained industrial environments.

Square-root Kalman filters propagate state covariances in Cholesky-factor form for numerical stability, and are a natural target for gradient-based parameter learning in state-space models. Their core operation, triangularization of a matrix $M \in \mathbb{R}^{n \times m}$, is computed via a QR decomposition in practice, but naively differentiating through it causes two problems: the semi-orthogonal factor is non-unique when $m > n$, yielding undefined gradients; and the standard Jacobian formula involves inverses, which diverges when $M$ is rank-deficient. Both are resolved by the observation that all filter outputs relevant to learning depend on the input matrix only through the Gramian $MM^\top$, so the composite loss is smooth in $M$ even where the triangularization is not. We derive a closed-form chain-rule directly from the differential of this Gramian identity, prove it exact for the Kalman log-marginal likelihood and filtered moments, and extend it to rank-deficient inputs via a two-component decomposition: a column-space term based on the Moore--Penrose pseudoinverse, and a null-space correction for perturbations outside the column space of $M$.

We introduce Holographic Invariant Storage (HIS), a protocol that assembles known properties of bipolar Vector Symbolic Architectures into a design-time safety contract for LLM context-drift mitigation. The contract provides three closed-form guarantees evaluable before deployment: single-signal recovery fidelity converging to $1/\sqrt{2} \approx 0.707$ (regardless of noise depth or content), continuous-noise robustness $2Φ(1/σ) - 1$, and multi-signal capacity degradation $\approx\sqrt{1/(K+1)}$. These bounds, validated by Monte Carlo simulation ($n = 1{,}000$), enable a systems engineer to budget recovery fidelity and codebook capacity at design time -- a property no timer or embedding-distance metric provides. A pilot behavioral experiment (four LLMs, 2B--7B, 720 trials) confirms that safety re-injection improves adherence at the 2B scale; full results are in an appendix.

Feature matching is a fundamental problem in computer vision with wide-ranging applications, including simultaneous localization and mapping (SLAM), image stitching, and 3D reconstruction. While recent advances in deep learning have improved keypoint detection and description, most approaches focus primarily on geometric attributes and often neglect higher-level semantic information. This work proposes a semantic-aware feature extraction framework that employs multi-task learning to jointly train keypoint detection, keypoint description, and semantic segmentation. The method is benchmarked against standard feature matching techniques and evaluated in the context of 3D reconstruction. To enhance feature correspondence, a deep matching module is integrated. The system is tested using input from a single monocular fisheye camera mounted on a vehicle and evaluated within a multi-floor parking structure. The proposed approach supports semantic 3D reconstruction with altitude estimation, capturing elevation changes and enabling multi-level mapping. Experimental results demonstrate that the method produces semantically annotated 3D point clouds with improved structural detail and elevation information, underscoring the effectiveness of joint training with semantic cues for more consistent feature matching and enhanced 3D reconstruction.

We present a scalable formal verification methodology for Quantum Phase Estimation (QPE) circuits. Our approach uses a symbolic qubit abstraction based on quantifier-free bit-vector logic, capturing key quantum phenomena, including superposition, rotation, and measurement. The proposed methodology maps quantum circuit functional behaviour from Hilbert space to a bit-vector domain. We develop formal properties aligned with this abstraction to ensure functional correctness of QPE circuits. The method scales efficiently, verifying QPE circuits with up to 6 precision qubits and 1,024 phase qubits using under 3.5 GB of memory.

Optimization analyses for cross-entropy training rely on local Taylor models of the loss to predict whether a proposed step will decrease the objective. These surrogates are reliable only inside the Taylor convergence radius of the true loss along the update direction. That radius is set not by real-line curvature alone but by the nearest complex singularity. For cross-entropy, the softmax partition function $F=\sum_j \exp(z_j)$ has complex zeros -- ``ghosts of softmax'' -- that induce logarithmic singularities in the loss and cap this radius. To make this geometry usable, we derive closed-form expressions under logit linearization along the proposed update direction. In the binary case, the exact radius is $ρ^*=\sqrt{δ^2+ π^2}/Δ_a$. In the multiclass case, we obtain the lower bound $ρ_a=π/Δ_a$, where $Δ_a=\max_k a_k-\min_k a_k$ is the spread of directional logit derivatives $a_k=\nabla z_k\cdot v$. This bound costs one Jacobian-vector product and reveals what makes a step fragile: samples that are both near a decision flip and highly sensitive to the proposed direction tighten the radius. The normalized step size $r=τ/ρ_a$ separates safe from dangerous updates. Across six tested architectures and multiple step directions, no model fails for $r<1$, yet collapse appears once $r\ge 1$. Temperature scaling confirms the mechanism: normalizing by $ρ_a$ shrinks the onset-threshold spread from standard deviation $0.992$ to $0.164$. A controller that enforces $τ\leρ_a$ survives learning-rate spikes up to $10{,} 000\times$ in our tests, where gradient clipping still collapses. Together, these results identify a geometric constraint on cross-entropy optimization that operates through Taylor convergence rather than Hessian curvature.

We introduce Probabilistic Gaussian Homotopy (PGH), a probability-space continuation framework for nonconvex optimization. Unlike classical Gaussian homotopy, which smooths the objective and uniformly averages gradients, PGH deforms the associated Boltzmann distribution and induces Boltzmann-weighted aggregation of perturbed gradients, which exponentially biases descent directions toward low-energy regions. We show that PGH corresponds to a log-sum-exp (soft-min) homotopy that smooths a nonconvex objective at scale $λ>0$ and recovers the original objective as $λ\to 0$, yielding a posterior-mean generalization of the Moreau envelope, and we derive a dynamical system governing minimizer evolution along an annealed homotopy path. This establishes a principled connection between Gaussian continuation, Bayesian denoising, and diffusion-style smoothing. We further propose Probabilistic Gaussian Homotopy Optimization (PGHO), a practical stochastic algorithm based on Monte Carlo gradient estimation, and demonstrate strong performance on high-dimensional nonconvex benchmarks and sparse recovery problems where classical gradient methods and objective-space smoothing frequently fail.

AI development has a fiction dependency problem: models are built on massive corpora of modern fiction and desperately need more of it, yet they struggle to generate it. I term this the AI-Fiction Paradox and it is particularly startling because in machine learning, training data typically determines output quality. This paper offers a theoretically precise account of why fiction resists AI generation by identifying three distinct challenges for current architectures. First, fiction depends on what I call narrative causation, a form of plot logic where events must feel both surprising in the moment and retrospectively inevitable. This temporal paradox fundamentally conflicts with the forward-generation logic of transformer architectures. Second, I identify an informational revaluation challenge: fiction systematically violates the computational assumption that informational importance aligns with statistical salience, requiring readers and models alike to retrospectively reweight the significance of narrative details in ways that current attention mechanisms cannot perform. Third, drawing on over seven years of collaborative research on sentiment arcs, I argue that compelling fiction requires multi-scale emotional architecture, the orchestration of sentiment at word, sentence, scene, and arc levels simultaneously. Together, these three challenges explain both why AI companies have risked billion-dollar lawsuits for access to modern fiction and why that fiction remains so difficult to replicate. The analysis also raises urgent questions about what happens when these challenges are overcome. Fiction concentrates uniquely powerful cognitive and emotional patterns for modeling human behavior, and mastery of these patterns by AI systems would represent not just a creative achievement but a potent vehicle for human manipulation at scale.

The use of ensemble-based multi-label methods has been shown to be effective in improving multi-label classification results. One of the most widely used ensemble-based multi-label classifiers is Ensemble of Classifier Chains. Decision templates for Ensemble of Classifier Chains (DTECC) is a fusion scheme based on Decision Templates that combines the predictions of Ensemble of Classifier Chains using information from the decision profile for each label, without considering information about other labels that might contribute to the classified result. Based on DTECC, this work proposes the Unconditionally Dependent Decision Templates for Ensemble of Classifier Chains (UDDTECC) method, a classifier fusion method that seeks to exploit correlations between labels in the fusion process. In this way, the classification of each label in the problem takes into account the label values that are considered conditionally dependent and that can lead to an improvement in the classification performance. The proposed method is experimentally compared with two traditional classifier fusion strategies and with a stacking-based strategy. Empirical evidence shows that using the proposed Decision Templates adaptation can improve the performance compared to the traditionally used fusion schemes on most of the evaluated metrics.

We present AMES (Approximate Multimodal Enterprise Search), a unified multimodal late interaction retrieval architecture which is backend agnostic. AMES demonstrates that fine-grained multimodal late interaction retrieval can be deployed within a production grade enterprise search engine without architectural redesign. Text tokens, image patches, and video frames are embedded into a shared representation space using multi-vector encoders, enabling cross-modal retrieval without modality specific retrieval logic. AMES employs a two-stage pipeline: parallel token level ANN search with per document Top-M MaxSim approximation, followed by accelerator optimized Exact MaxSim re-ranking. Experiments on the ViDoRe V3 benchmark show that AMES achieves competitive ranking performance within a scalable, production ready Solr based system.

Near-term quantum devices provide only finite-shot measurements and prepare imperfect, contaminated states. This motivates algorithms that convert samples into reliable low-energy estimates without full tomography or exhaustive measurements. We propose Active Sampling Sample-based Quantum Diagonalization (AS-SQD), framing SQD as an active learning problem: given measured bitstrings, which additional basis states should be included to efficiently recover the ground-state energy? SQD restricts the Hamiltonian to a selected set of basis states and classically diagonalizes the restricted matrix. However, naive SQD using only sampled states suffers from bias under finite-shot sampling and excited-state contamination, while blind random expansion is inefficient as system size grows. We introduce a perturbation-theoretic acquisition function based on Epstein--Nesbet second-order energy corrections to rank candidate basis states connected to the current subspace. At each iteration, AS-SQD diagonalizes the restricted Hamiltonian, generates connected candidates, and adds the most valuable ones according to this score. We evaluate AS-SQD on disordered Heisenberg and Transverse-Field Ising (TFIM) spin chains up to 16 qubits under a preparation model mixing 80\% ground state and 20\% first excited state. Furthermore, we validate its robustness against real-world state preparation and measurement (SPAM) errors using physical samples from an IBM Quantum processor. Across simulated and hardware evaluations, AS-SQD consistently achieves substantially lower absolute energy errors than standard SQD and random expansion. Detailed ablation studies demonstrate that physics-guided basis acquisition effectively concentrates computation on energetically relevant directions, bypassing exponential combinatorial bottlenecks.

Coordinated operations of multi-robot systems (MRS) require agents to maintain communication connections to accomplish team objectives. However, maintaining the connections imposes costs in terms of restricted robot mobility, resulting in suboptimal team performance. In this work, we consider a realistic MRS framework in which agents are subject to unknown dynamical disturbances and experience communication delays. Most existing works on connectivity maintenance use consensus-based frameworks for graph reconfiguration, where decision-making time scales with the number of nodes and requires multiple rounds of communication, making them ineffective under communication delays. To address this, we propose a novel leader-based decision-making algorithm that uses a central node for efficient real-time reconfiguration, reducing decision-making time to depend on the graph diameter rather than the number of nodes and requiring only one round of information transfer through the network. We propose a novel method for estimating robot locations within the MRS that actively accounts for unknown disturbances and the communication delays. Using these position estimates, the central node selects a set of edges to delete while allowing the formation of new edges, aiming to keep the diameter of the new graph within a threshold. We provide numerous simulation results to showcase the efficacy of the proposed method.

The growing availability of Earth Observation (EO) data and recent advances in Computer Vision have driven rapid progress in machine learning for EO, producing domain-specific models at ever-increasing scales. Yet this progress risks overlooking fundamental properties of EO data that distinguish it from other domains. We argue that EO data exhibit a multidimensional redundancy (spectral, temporal, spatial, and semantic) which has a more pronounced impact on the domain and its applications than what current literature reflects. To validate this hypothesis, we conduct a systematic domain-specific investigation examining the existence, consistency, and practical implications of this phenomenon across key dimensions of EO variability. Our findings confirm that redundancy in EO data is both substantial and pervasive: exploiting it yields comparable performance ($\approx98.5\%$ of baseline) at a fraction of the computational cost ($\approx4\times$ fewer GFLOPs), at both training and inference. Crucially, these gains are consistent across tasks, geospatial locations, sensors, ground sampling distances, and architectural designs; suggesting that multi-faceted redundancy is a structural property of EO data rather than an artifact of specific experimental choices. These results lay the groundwork for more efficient, scalable, and accessible large-scale EO models.

Full-stream text-to-speech (TTS) for interactive systems must start speaking with minimal delay while remaining controllable as text arrives incrementally. We present VoXtream2, a zero-shot full-stream TTS model with dynamic speaking-rate control that can be updated mid-utterance on the fly. VoXtream2 combines a distribution matching mechanism over duration states with classifier-free guidance across conditioning signals to improve controllability and synthesis quality. Prompt-text masking enables textless audio prompting, removing the need for prompt transcription. Across standard zero-shot benchmarks and a dedicated speaking-rate test set, VoXtream2 achieves competitive objective and subjective results against public baselines despite a smaller model and less training data. In full-stream mode, it runs 4 times faster than real time with 74 ms first-packet latency on a consumer GPU.

The Logic Theorist (LT), created by Allen Newell, J. C. Shaw, and Herbert Simon in 1955-1956, is widely regarded as the first artificial intelligence program. While the original conceptual model was described in 1956, it underwent several iterations as the underlying Information Processing Language (IPL) evolved. Here I describe the construction of a new IPL-V interpreter, written in Common Lisp, and the faithful reanimation of the Logic Theorist from code transcribed directly from Stefferud's 1963 RAND technical report. Stefferud's version represents a pedagogical re-coding of the original heuristic logic into the standardized IPL-V. The reanimated LT successfully proves 16 of 23 attempted theorems from Chapter 2 of Principia Mathematica, results that are historically consistent with the original system's behavior within its search limits. To the author's knowledge, this is the first successful execution of the original Logic Theorist code in over half a century.

Asynchronous Bayesian optimization is widely used for gradient-free optimization in domains with independent parallel experiments and varying evaluation times. Existing methods posit that standard acquisitions lead to redundant and repeated queries, proposing complex solutions to enforce diversity in queries. Challenging this fundamental premise, we show that methods, like the Upper Confidence Bound, can in fact achieve theoretical guarantees essentially equivalent to those of sequential Thompson sampling. A conceptual analysis of asynchronous Bayesian optimization reveals that existing works neglect intermediate posterior updates, which we find to be generally sufficient to avoid redundant queries. Further investigation shows that by penalizing busy locations, diversity-enforcing methods can over-explore in asynchronous settings, reducing their performance. Our extensive experiments demonstrate that simple standard acquisition functions match or outperform purpose-built asynchronous methods across synthetic and real-world tasks.

Melanoma is the most lethal form of skin cancer, and early detection is critical for improving patient outcomes. Although dermoscopy combined with deep learning has advanced automated skin-lesion analysis, progress is hindered by limited access to large, well-annotated datasets and by severe class imbalance, where melanoma images are substantially underrepresented. To address these challenges, we present the first systematic benchmarking study comparing four GAN architectures-DCGAN, StyleGAN2, and two StyleGAN3 variants (T/R)-for high-resolution melanoma-specific synthesis. We train and optimize all models on two expert-annotated benchmarks (ISIC 2018 and ISIC 2020) under unified preprocessing and hyperparameter exploration, with particular attention to R1 regularization tuning. Image quality is assessed through a multi-faceted protocol combining distribution-level metrics (FID), sample-level representativeness (FMD), qualitative dermoscopic inspection, downstream classification with a frozen EfficientNet-based melanoma detector, and independent evaluation by two board-certified dermatologists. StyleGAN2 achieves the best balance of quantitative performance and perceptual quality, attaining FID scores of 24.8 (ISIC 2018) and 7.96 (ISIC 2020) at gamma=0.8. The frozen classifier recognizes 83% of StyleGAN2-generated images as melanoma, while dermatologists distinguish synthetic from real images at only 66.5% accuracy (chance = 50%), with low inter-rater agreement (kappa = 0.17). In a controlled augmentation experiment, adding synthetic melanoma images to address class imbalance improved melanoma detection AUC from 0.925 to 0.945 on a held-out real-image test set. These findings demonstrate that StyleGAN2-generated melanoma images preserve diagnostically relevant features and can provide a measurable benefit for mitigating class imbalance in melanoma-focused machine learning pipelines.

Constructing reduced-order models (ROMs) capable of efficiently predicting the evolution of high-dimensional, parametric systems is crucial in many applications in engineering and applied sciences. A popular class of projection-based ROMs projects the high-dimensional full-order model (FOM) dynamics onto a low-dimensional manifold. These projection-based ROMs approaches often rely on classical model reduction techniques such as proper orthogonal decomposition (POD) or, more recently, on neural network architectures such as autoencoders (AEs). In the case that the ROM is constructed by the POD, one has approximation guaranteed based based on the singular values of the problem at hand. However, POD-based techniques can suffer from slow decay of the singular values in transport- and advection-dominated problems. In contrast to that, AEs allow for better reduction capabilities than the POD, often with the first few modes, but at the price of theoretical considerations. In addition, it is often observed, that AEs exhibits a plateau of the projection error with the increment of the dimension of the trial manifold. In this work, we propose a deep invertible AE architecture, named inv-AE, that improves upon the stagnation of the projection error typical of traditional AE architectures, e.g., convolutional, and the reconstructions quality. Inv-AE is composed of several invertible neural network layers that allows for gradually recovering more information about the FOM solutions the more we increase the dimension of the reduced manifold. Through the application of inv-AE to a parametric 1D Burgers' equation and a parametric 2D fluid flow around an obstacle with variable geometry, we show that (i) inv-AE mitigates the issue of the characteristic plateau of (convolutional) AEs and (ii) inv-AE can be combined with popular projection-based ROM approaches to improve their accuracy.

In a spiking neural network, is it enough for each neuron to spike at most once? In recent work, approximation bounds for spiking neural networks have been derived, quantifying how well they can fit target functions. However, these results are only valid for neurons that spike at most once, which is commonly thought to be a strong limitation. Here, we show that the opposite is true for a large class of spiking neuron models, including the commonly used leaky integrate-and-fire model with subtractive reset: for every approximation bound that is valid for a set of multi-spike neural networks, there is an equivalent set of single-spike neural networks with only linearly more neurons (in the maximum number of spikes) for which the bound holds. The same is true for the reverse direction too, showing that regarding their approximation capabilities in general machine learning tasks, single-spike and multi-spike neural networks are equivalent. Consequently, many approximation results in the literature for single-spike neural networks also hold for the multi-spike case.

Model merging has shown that multitask models can be created by directly combining the parameters of different models that are each specialized on tasks of interest. However, models trained independently on distinct tasks often exhibit interference that degrades the merged model's performance. To solve this problem, we formally define the notion of Cross-Task Interference as the drift in the representation of the merged model relative to its constituent models. Reducing cross-task interference is key to improving merging performance. To address this issue, we propose our method, Resolving Interference (RI), a light-weight adaptation framework which disentangles expert models to be functionally orthogonal to the space of other tasks, thereby reducing cross-task interference. RI does this whilst using only unlabeled auxiliary data as input (i.e., no task-data is needed), allowing it to be applied in data-scarce scenarios. RI consistently improves the performance of state-of-the-art merging methods by up to 3.8% and generalization to unseen domains by up to 2.3%. We also find RI to be robust to the source of auxiliary input while being significantly less sensitive to tuning of merging hyperparameters. Our codebase is available at: https://github.com/pramesh39/resolving_interference

Backdoor attacks pose severe security threats to large language models (LLMs), where a model behaves normally under benign inputs but produces malicious outputs when a hidden trigger appears. Existing backdoor removal methods typically assume prior knowledge of triggers, access to a clean reference model, or rely on aggressive finetuning configurations, and are often limited to classification tasks. However, such assumptions fall apart in real-world instruction-tuned LLM settings. In this work, we propose a new framework for purifying instruction-tuned LLM without any prior trigger knowledge or clean references. Through systematic sanity checks, we find that backdoor associations are redundantly encoded across MLP layers, while attention modules primarily amplify trigger signals without establishing the behavior. Leveraging this insight, we shift the focus from isolating specific backdoor triggers to cutting off the trigger-behavior associations, and design an immunization-inspired elimination approach: by constructing multiple synthetic backdoored variants of the given suspicious model, each trained with different malicious trigger-behavior pairs, and contrasting them with their clean counterparts. The recurring modifications across variants reveal a shared "backdoor signature"-analogous to antigens in a virus. Guided by this signature, we neutralize highly suspicious components in LLM and apply lightweight finetuning to restore its fluency, producing purified models that withstand diverse backdoor attacks and threat models while preserving generative capability.

Large Language Models (LLMs) increasingly serve as autonomous reasoning agents in decision support, scientific problem-solving, and multi-agent coordination systems. However, deploying LLM agents in consequential applications requires assurance that their reasoning remains stable under semantically equivalent input variations, a property we term semantic invariance. Standard benchmark evaluations, which assess accuracy on fixed, canonical problem formulations, fail to capture this critical reliability dimension. To address this shortcoming, in this paper we present a metamorphic testing framework for systematically assessing the robustness of LLM reasoning agents, applying eight semantic-preserving transformations (identity, paraphrase, fact reordering, expansion, contraction, academic context, business context, and contrastive formulation) across seven foundation models spanning four distinct architectural families: Hermes (70B, 405B), Qwen3 (30B-A3B, 235B-A22B), DeepSeek-R1, and gpt-oss (20B, 120B). Our evaluation encompasses 19 multi-step reasoning problems across eight scientific domains. The results reveal that model scale does not predict robustness: the smaller Qwen3-30B-A3B achieves the highest stability (79.6% invariant responses, semantic similarity 0.91), while larger models exhibit greater fragility.

In-context learning (ICL) is a valuable capability exhibited by Transformers pretrained on diverse sequence tasks. However, previous studies have observed that ICL often conflicts with the model's inherent in-weight learning (IWL) ability. By examining the representation space learned by a toy model in synthetic experiments, we identify the shared encoding space for context and samples in Transformers as a potential source of this conflict. To address this, we modify the model architecture to separately encode the context and samples into two distinct spaces: a task representation space and a sample representation space. We model these two spaces under a simple yet principled framework, assuming a linear representational structure and treating them as a pair of dual spaces. Both theoretical analysis and empirical results demonstrate the effectiveness of our proposed architecture, CoQE, in the single-value answer setting. It not only enhances ICL performance through improved representation learning, but also successfully reconciles ICL and IWL capabilities across synthetic few-shot classification and a newly designed pseudo-arithmetic task. Code: https://github.com/McGuinnessChen/dual-representation-space-encoding

Drawing on recent breakthroughs in cellular neurobiology and detailed biophysical modeling linking neocortical pyramidal neurons to distinct mental-state regimes, this work introduces a mathematically grounded formulation showing how models (e.g., Transformers) can implement computational principles underlying awake imaginative thought to pre-select relevant information before attention is applied via triadic modulation loops among queries ($Q$), keys ($K$), and values ($V$).~Scalability experiments on ImageNet-1K, benchmarked against a standard Vision Transformer (ViT), demonstrate significantly faster learning with reduced computational demand (fewer heads, layers, and tokens), consistent with our prior findings in reinforcement learning and language modeling. The approach operates at approximately $\mathcal{O}(N)$ complexity with respect to the number of input tokens $N$.

We introduce CRYSTAL (Clear Reasoning via Yielded Steps, Traceability, and Logic), a diagnostic benchmark with 6,372 instances that evaluates multimodal reasoning through verifiable intermediate steps. We propose two complementary metrics: Match F1, which scores step-level precision and recall via semantic similarity matching, and Ordered Match F1, which further penalizes disordered reasoning chains. References are constructed through a Delphi-inspired pipeline in which four independent MLLMs generate trajectories, which are then aggregated via semantic clustering and validated through human quality gates. Evaluation of 20 MLLMs, including commercial frontier systems not used during benchmark construction, reveals systematic failures that are invisible to answer accuracy: universal cherry-picking (precision far exceeds recall), non-monotonic scaling trade-offs, and disordered reasoning in which no competitive model preserves more than 60% of matched steps in the correct order. Beyond evaluation, we propose the Causal Process Reward (CPR), a multiplicative reward that couples answer correctness with step-level alignment, and CPR-Curriculum, which progressively increases reasoning difficulty during training. CPR-Curriculum achieves a 32% improvement in Match F1 via GRPO where additive reward strategies fail, improving reasoning without manual step annotation.

Research about bias in machine learning has mostly focused on outcome-oriented fairness metrics (e.g., equalized odds) and on a single protected category. Although these approaches offer great insight into bias in ML, they provide limited insight into model procedure bias. To address this gap, we proposed multi-category explanation stability disparity (MESD), an intersectional, procedurally oriented metric that measures the disparity in the quality of explanations across intersectional subgroups in multiple protected categories. MESD serves as a complementary metric to outcome-oriented metrics, providing detailed insight into the procedure of a model. To further extend the scope of the holistic selection model, we also propose a multi-objective optimization framework, UEF (Utility-Explanation-Fairness), that jointly optimizes three objectives. Experimental results across multiple datasets show that UEF effectively balances objectives. Also, the results show that MESD can effectively capture the explanation difference between intersectional groups. This research addresses an important gap by examining explainability with respect to fairness across multiple protected categories.

Discrete Diffusion Language Models have emerged as a compelling paradigm for unified multimodal generation, yet their deployment is hindered by high inference latency arising from iterative decoding. Existing acceleration strategies often require expensive re-training or fail to leverage the 2D spatial redundancy inherent in visual data. To address this, we propose Locality-Aware Dynamic Rescue (LADR), a training-free method that expedites inference by exploiting the spatial Markov property of images. LADR prioritizes the recovery of tokens at the ''generation frontier'', regions spatially adjacent to observed pixels, thereby maximizing information gain. Specifically, our method integrates morphological neighbor identification to locate candidate tokens, employs a risk-bounded filtering mechanism to prevent error propagation, and utilizes manifold-consistent inverse scheduling to align the diffusion trajectory with the accelerated mask density. Extensive experiments on four text-to-image generation benchmarks demonstrate that our LADR achieves an approximate 4 x speedup over standard baselines. Remarkably, it maintains or even enhances generative fidelity, particularly in spatial reasoning tasks, offering a state-of-the-art trade-off between efficiency and quality.

Tropical cyclone (TC) intensity forecasting remains challenging as current numerical and AI-based weather models fail to satisfactorily represent extreme TC structure and intensity. Although intensity time-series forecasting has achieved significant advances, it outputs intensity sequences rather than the three-dimensional inner-core fine-scale structure and physical mechanisms governing TC evolution. High-resolution numerical simulations can capture these features but remain computationally expensive and inefficient for large-scale operational applications. Here we present 3DTCR, a physics-based generative framework combining physical constraints with generative AI efficiency for 3D TC structure reconstruction. Trained on a six-year, 3-km-resolution moving-domain WRF dataset, 3DTCR enables region-adaptive vortex-following reconstruction using conditional Flow Matching(CFM), optimized via latent domain adaptation and two-stage transfer learning. The framework mitigates limitations imposed by low-resolution targets and over-smoothed forecasts, improving the representation of TC inner-core structure and intensity while maintaining track stability. Results demonstrate that 3DTCR outperforms the ECMWF high-resolution forecasting system (ECMWF-HRES) in TC intensity prediction at nearly all lead times up to 5 days and reduces the RMSE of maximum WS10M by 36.5% relative to its FuXi inputs. These findings highlight 3DTCR as a physics-based generative framework that efficiently resolves fine-scale structures at lower computational cost, which may offer a promising avenue for improving TC intensity forecasting.

An X-ray computed tomography (CT), metal artifact reduction (MAR) remains a major challenge because metallic implants violate standard CT forward-model assumptions, producing severe streaking and shadowing artifacts that degrade diagnostic quality. We propose MGMAR, a metal-guided MAR method that explicitly leverages metal-related information throughout the reconstruction pipeline. MGMAR first generates a high-quality prior image by training a conditioned implicit neural representation (INR) using metal-unaffected projections, and then incorporates this prior into a normalized MAR (NMAR) framework for projection completion. To improve robustness under severe metal corruption, we pretrain the encoder-conditioned INR on paired metal-corrupted and artifact-free CT images, thereby embedding data-driven prior knowledge into the INR parameter space. This prior-embedded initialization reduces sensitivity to random initialization and accelerates convergence during measurement-specific refinement. The encoder takes a metal-corrupted reconstruction together with a recursively constructed metal artifact image, enabling the latent field to capture metal-dependent global artifact patterns. After projection completion using the INR prior, we further suppress residual artifacts using a metal-conditioned correction network, where the metal mask modulates intermediate features via adaptive instance normalization to target metal-dependent secondary artifacts while preserving anatomical structures. Experiments on the public AAPM-MAR benchmark demonstrate that MGMAR achieves state-of-the-art performance, attaining an average final score of 0.89 on 29 clinical test cases.

Predicting atomic-scale crack propagation in aluminum nitride (AlN) is critical for semiconductor reliability but remains prohibitively expensive via molecular dynamics (MD). We develop a diffusion-based generative machine learning model to predict atomic-scale crack propagation in AlN, a critical semiconductor material, by conditioning solely on initial microstructure embeddings. Trained on MD simulations of single-crack systems, the model achieves a significant speedup while accurately forecasting dynamic fracture processes, including stress-driven crack initiation, crack branching, and atomic-scale bridging ligaments. Crucially, it demonstrates inherent physical fidelity by reproducing material-intrinsic mechanisms while disregarding periodic boundary artifacts, and generalizes to unseen multi-crack configurations. Validation against MD ground truth confirms the capability of the model to capture complex fracture physics without auxiliary stress or energy data, enabling rapid exploration of crack-mediated failure for semiconductor reliability optimization.

We present a technical case study on the Privacy-Enhancing Technologies (PETs) for Public Health Challenge, a collaborative effort to safely leverage sensitive private sector data for social impact, specifically pandemic management. The project utilized Differential Privacy (DP) to create realistic, privacy-preserved synthetic financial transaction data, which was then combined with public health and mobility datasets. This approach successfully addressed the critical hurdle of sharing sensitive financial information for research and policy. The analysis demonstrated that this synthetic, DP-protected data possesses significant spatial-temporal and predictive power for public health. Key outcomes include the development of six reusable tools and frameworks supporting diagnostic nowcasting (e.g., Hotspot Detection, Pandemic Adherence Monitoring) and predictive forecasting (e.g., Mobility Analysis, Contact Matrix Estimation) for epidemiological decision-making. The study provides best practices for advancing data sharing in a privacy-compliant manner.

We present NormCode Canvas (v1.1.3), a deployed system realizing Case-Based Reasoning at two levels for multi-step LLM workflows. The foundation is NormCode, a semi-formal planning language whose compiler-verified scope rule ensures every execution checkpoint is a genuinely self-contained case -- eliminating the implicit shared state that makes retrieval unreliable and failure non-localizable in standard orchestration frameworks. Level 1 treats each checkpoint as a concrete case (suspended runtime); Fork implements retrieve-and-reuse, Value Override implements revision with automatic stale-boundary propagation. Level 2 treats each compiled plan as an abstract case; the compilation pipeline is itself a NormCode plan, enabling recursive case learning. Three structural properties follow: (C1) direct checkpoint inspection; (C2) pre-execution review via compiler-generated narrative; (C3) scope-bounded selective re-execution. Four deployed plans serve as structured evidence: PPT Generation produces presentation decks at ~40s per slide on commercial APIs; Code Assistant carries out multi-step software-engineering tasks spanning up to ten reasoning cycles; NC Compilations converts natural-language specifications into executable NormCode plans; and Canvas Assistant, when connected to an external AI code editor, automates plan debugging. Together these plans form a self-sustaining ecosystem in which plans produce, debug, and refine one another -- realizing cumulative case-based learning at system scale.

Adapting Large Language Models (LLMs) to specialized domains requires high-quality instruction tuning datasets, which are expensive to create through human annotation. Existing data synthesis methods focus on general-purpose tasks and fail to capture domain-specific terminology and reasoning patterns. To address this, we introduce DS$^2$-Instruct, a zero-shot framework that generates domain-specific instruction datasets without human supervision. Our approach first generates task-informed keywords to ensure comprehensive domain coverage. It then creates diverse instructions by pairing these keywords with different cognitive levels from Bloom's Taxonomy. Finally, it uses self-consistency validation to ensure data quality. We apply this framework to generate datasets across seven challenging domains, such as mathematics, finance, and logical reasoning. Comprehensive evaluation demonstrates that models fine-tuned on our generated data achieve substantial improvements over existing data generation methods.

Quantum principal component analysis (qPCA) is commonly formulated as the extraction of eigenvalues and eigenvectors of a covariance-encoded density operator. Yet in many qPCA settings, the practical objective is simpler: projecting data onto the dominant spectral subspace. In this work, we introduce a projection-first framework, the Filtered Spectral Projection Algorithm (FSPA), which bypasses explicit eigenvalue estimation while preserving the essential spectral structure. FSPA amplifies any nonzero warm-start overlap with the leading principal subspace and remains robust in small-gap and near-degenerate regimes without inducing artificial symmetry breaking in the absence of bias. To connect this approach to classical datasets, we show that for amplitude-encoded centered data, the ensemble density matrix $ρ=\sum_i p_i|ψ_i\rangle\langleψ_i|$ coincides with the covariance matrix. For uncentered data, $ρ$ corresponds to PCA without centering, and we derive eigenvalue interlacing bounds quantifying the deviation from standard PCA. We further show that ensembles of quantum states admit an equivalent centered covariance interpretation. Numerical demonstrations on benchmark datasets, including Breast Cancer Wisconsin and handwritten Digits, show that downstream performance remains stable whenever projection quality is preserved. These results suggest that, in a broad class of qPCA settings, spectral projection is the essential primitive, and explicit eigenvalue estimation is often unnecessary.

Deep generative models offer a powerful alternative to conventional channel estimation by learning complex channel distributions. By integrating the rich environmental information available in modern sensing-aided networks, this paper proposes MultiCE-Flow, a multimodal channel estimation framework based on flow matching and diffusion transformer (DiT). We design a specialized multimodal perception module that fuses LiDAR, camera, and location data into a semantic condition, while treating sparse pilots as a structural condition. These conditions guide a DiT backbone to reconstruct high-fidelity channels. Unlike standard diffusion models, we employ flow matching to learn a linear trajectory from noise to data, enabling efficient one-step sampling. By leveraging environmental semantics, our method mitigates the ill-posed nature of estimation with sparse pilots. Extensive experiments demonstrate that MultiCE-Flow consistently outperforms traditional baselines and existing generative models. Notably, it exhibits superior robustness to out-of-distribution scenarios and varying pilot densities, making it suitable for environment-aware communication systems.

Video generation models have been used as a robot policy to predict the future states of executing a task conditioned on task description and observation. Previous works ignore their high computational cost and long inference time. To address this challenge, we propose Draft-and-Target Sampling, a novel diffusion inference paradigm for video generation policy that is training-free and can improve inference efficiency. We introduce a self-play denoising approach by utilizing two complementary denoising trajectories in a single model, draft sampling takes large steps to generate a global trajectory in a fast manner and target sampling takes small steps to verify it. To further speedup generation, we introduce token chunking and progressive acceptance strategy to reduce redundant computation. Experiments on three benchmarks show that our method can achieve up to 2.1x speedup and improve the efficiency of current state-of-the-art methods with minimal compromise to the success rate. Our code is available.

Voice anonymization masks vocal traits while preserving linguistic content, which may still leak speaker-specific patterns. To assess and strengthen privacy evaluation, we propose a dual-stream attacker that fuses spectral and self-supervised learning features via parallel encoders with a three-stage training strategy. Stage I establishes foundational speaker-discriminative representations. Stage II leverages the shared identity-transformation characteristics of voice conversion and anonymization, exposing the model to diverse converted speech to build cross-system robustness. Stage III provides lightweight adaptation to target anonymized data. Results on the VoicePrivacy Attacker Challenge (VPAC) dataset demonstrate that Stage II is the primary driver of generalization, enabling strong attacking performance on unseen anonymization datasets. With Stage III, fine-tuning on only 10\% of the target anonymization dataset surpasses current state-of-the-art attackers in terms of EER.

Diffusion-based image-to-video (I2V) models increasingly exhibit world-model-like properties by implicitly capturing temporal dynamics. However, existing studies have mainly focused on visual quality and controllability, and the robustness of the state transition learned by the model remains understudied. To fill this gap, we are the first to analyze the vulnerability of I2V models, find that temporal control mechanisms constitute a new attack surface, and reveal the challenge of modeling them uniformly under different attack settings. Based on this, we propose a trajectory-control attack, called CtrlAttack, to interfere with state evolution during the generation process. Specifically, we represent the perturbation as a low-dimensional velocity field and construct a continuous displacement field via temporal integration, thereby affecting the model's state transitions while maintaining temporal consistency; meanwhile, we map the perturbation to the observation space, making the method applicable to both white-box and black-box attack settings. Experimental results show that even under low-dimensional and strongly regularized perturbation constraints, our method can still significantly disrupt temporal consistency by increasing the attack success rate (ASR) to over 90% in the white-box setting and over 80% in the black-box setting, while keeping the variation of the FID and FVD within 6 and 130, respectively, thus revealing the potential security risk of I2V models at the level of state dynamics.

In-context learning (ICL) converts static encoders into task-conditioned reasoners, enabling adaptation to new data from just a few examples without updating pretrained parameters. This capability is essential for graph foundation models (GFMs) to approach LLM-level generality. Yet current GFMs struggle with cross-domain alignment, typically relying on modality-specific encoders that fail when graphs are pre-vectorized or raw data is inaccessible. In this paper, we introduce Modality-Free Graph In-context Alignment (MF-GIA), a framework that makes a pretrained graph encoder promptable for few-shot prediction across heterogeneous domains without modality assumptions. MF-GIA captures domain characteristics through gradient fingerprints, which parameterize lightweight transformations that align pre-encoded features and indexed labels into unified semantic spaces. During pretraining, a dual prompt-aware attention mechanism with episodic objective learns to match queries against aligned support examples to establish prompt-based reasoning capabilities. At inference, MF-GIA performs parameter-update-free adaptation using only a few-shot support set to trigger cross-domain alignment and enable immediate prediction on unseen domains. Experiments demonstrate that MF-GIA achieves superior few-shot performance across diverse graph domains and strong generalization to unseen domains.

Knowledge editing (KE) aims to precisely rectify specific knowledge in Large Language Models (LLMs) without disrupting general capabilities. State-of-the-art methods suffer from an open-loop control mismatch. We identify a critical "Semantic-Execution Disconnect": the semantic target is derived independently without feedback from the downstream's feasible region. This misalignment often causes valid semantic targets to fall within the prohibited space, resulting in gradient truncation and editing failure. To bridge this gap, we propose MetaKE (Meta-learning Aligned Knowledge Editing), a new framework that reframes KE as a bi-level optimization problem. Departing from static calculation, MetaKE treats the edit target as a learnable meta-parameter: the upper-level optimizer seeks a feasible target to maximize post-edit performance, while the lower-level solver executes the editing. To address the challenge of differentiating through complex solvers, we derive a Structural Gradient Proxy, which explicitly backpropagates editability constraints to the target learning phase. Theoretical analysis demonstrates that MetaKE automatically aligns the edit direction with the model's feasible manifold. Extensive experiments confirm that MetaKE significantly outperforms strong baselines, offering a new perspective on knowledge editing.

Recent advances in robot manipulation increasingly leverage Vision-Language Models (VLMs) for high-level reasoning, such as decomposing task instructions into sequential action plans expressed in natural language that guide downstream low-level motor execution. However, current benchmarks do not assess whether these plans are spatially executable, particularly in specifying the exact spatial locations where the robot should interact to execute the plan, limiting evaluation of real-world manipulation capability. To bridge this gap, we define a novel task of grounded planning and introduce GroundedPlanBench, a newly curated benchmark for spatially grounded long-horizon action planning in the wild. GroundedPlanBench jointly evaluates hierarchical sub-action planning and spatial action grounding (where to act), enabling systematic assessment of whether generated sub-actions are spatially executable for robot manipulation. We further introduce Video-to-Spatially Grounded Planning (V2GP), an automated data generation framework that leverages real-world robot video demonstrations to improve spatially grounded long-horizon planning. Our evaluations reveal that spatially grounded long-horizon planning remains a major bottleneck for current VLMs. Our results demonstrate that V2GP provides a promising approach for improving both action planning and spatial grounding performance, validated on our benchmark as well as through real-world robot manipulation experiments, advancing progress toward spatially actionable planning.

Spatial Transcriptomics (ST) profiles thousands of gene expression values at discrete spots with precise coordinates on tissue sections, preserving spatial context essential for clinical and pathological studies. With rising sequencing throughput and advancing platforms, the expanding data volumes motivate large-scale ST pretraining. However, the fundamental unit for pretraining, i.e., what constitutes a single training sample, remains ill-posed. Existing choices fall into two camps: (1) treating each spot as an independent sample, which discards spatial dependencies and collapses ST into single-cell transcriptomics; and (2) treating an entire slide as a single sample, which produces prohibitively large inputs and drastically fewer training examples, undermining effective pretraining. To address this gap, we propose treating spatial transcriptomics as croppable images. Specifically, we define a multi-channel image representation with fixed spatial size by cropping patches from raw slides, thereby preserving spatial context while substantially increasing the number of training samples. Along the channel dimension, we define gene subset selection rules to control input dimensionality and improve pretraining stability. Extensive experiments show that the proposed image-like dataset construction for ST pretraining consistently improves downstream performance, outperforming conventional pretraining schemes. Ablation studies verify that both spatial patching and channel design are necessary, establishing a unified, practical paradigm for organizing ST data and enabling large-scale pretraining.

Computational antibody design has seen rapid methodological progress, with dozens of deep generative methods proposed in the past three years, yet the field lacks a standardized benchmark for fair comparison and model development. These methods are evaluated on different SAbDab snapshots, non-overlapping test sets, and incompatible metrics, and the literature fragments the design problem into numerous sub-tasks with no common definition. We introduce \textsc{Chimera-Bench} (\textbf{C}DR \textbf{M}odeling with \textbf{E}pitope-guided \textbf{R}edesign), a unified benchmark built around a single canonical task: \emph{epitope-conditioned CDR sequence-structure co-design}. \textsc{Chimera-Bench} provides (1) a curated, deduplicated dataset of \textbf{2,922} antibody-antigen complexes with epitope and paratope annotations; (2) three biologically motivated splits testing generalization to unseen epitopes, unseen antigen folds, and prospective temporal targets; and (3) a comprehensive evaluation protocol with five metric groups including novel epitope-specificity measures. We benchmark representative methods spanning different generative paradigms and report results across all splits. \textsc{Chimera-Bench} is the largest dataset of its kind for the antibody design problem, allowing the community to develop and test novel methods and evaluate their generalizability. The source code and data are available at: https://github.com/mansoor181/chimera-bench.git

Dynamic sparse attention (DSA) reduces the per-token attention bandwidth by restricting computation to a top-k subset of cached key-value (KV) entries, but its token-dependent selection pattern introduces a system-level challenge: the KV working set is fragmented, volatile, and difficult to prefetch, which can translate into poor cache locality and stalled decode throughput. We study these effects by implementing a lightweight indexer for DSA-style selection on multiple open-source backbones and logging per-layer KV indices during autoregressive decoding. Our analysis shows a gap in serving DSA backbones - a potential for a high volume of blocking LL (last level) cache miss events, causing inefficiency; we propose a novel LL cache reservation system to save KV tokens in the LL cache between decode steps, combined with a token-granularity LRU eviction policy, and show on the data we collected how this architecture can benefit serving with DSA implemented on different backbones. Finally, we propose directions for future architectural and algorithmic exploration to improve serving of DSA on modern inference platforms.

With AI agents increasingly deployed as long-running systems, it becomes essential to autonomously construct and continuously evolve customized software to enable interaction within dynamic environments. Yet, existing benchmarks evaluate agents on isolated, one-off coding tasks, neglecting the temporal dependencies and technical debt inherent in real-world software evolution. To bridge this gap, we introduce DeepCommit, an agentic pipeline that reconstructs verifiable Milestone DAGs from noisy commit logs, where milestones are defined as semantically cohesive development goals. These executable sequences enable EvoClaw, a novel benchmark that requires agents to sustain system integrity and limit error accumulation, dimensions of long-term software evolution largely missing from current benchmarks. Our evaluation of 12 frontier models across 4 agent frameworks reveals a critical vulnerability: overall performance scores drop significantly from $>$80% on isolated tasks to at most 38% in continuous settings, exposing agents' profound struggle with long-term maintenance and error propagation.

In different multimodal scenarios, it needs to integrate and utilize information across modalities in a specific way based on the demands of the task. Different integration ways between modalities are referred to as "multimodal interaction". How well a model handles various multimodal interactions largely characterizes its multimodal ability. In this paper, we introduce MIBench, a comprehensive benchmark designed to evaluate the multimodal interaction capabilities of Large Multimodal Models (LMMs), which formulates each instance as a (con_v , con_t, task) triplet with contexts from vision and text, necessitating that LMMs employ correct forms of multimodal interaction to effectively complete the task. MIBench assesses models from three key aspects: the ability to source information from vision-centric or text-centric cues, and the ability to generate new information from their joint synergy. Each interaction capability is evaluated hierarchically across three cognitive levels: Recognition, Understanding, and Reasoning. MIBench comprises over 10,000 vision-text context pairs spanning 32 distinct tasks. Evaluation of state-of-the-art LMMs show that: (1) LMMs' ability on multimodal interaction remains constrained, despite the scaling of model parameters and training data; (2) they are easily distracted by textual modalities when processing vision information; (3) they mostly possess a basic capacity for multimodal synergy; and (4) natively trained multimodal models show noticeable deficits in fundamental interaction ability. We expect that these observations can serve as a reference for developing LMMs with more enhanced multimodal ability in the future.

Semantic routers in LLM inference gateways select tools in the critical request path, where every millisecond of added latency compounds across millions of requests. We propose Outcome-Aware Tool Selection (OATS), which interpolates tool embeddings toward the centroid of queries where they historically succeed -- an offline process that adds no parameters, latency, or GPU cost at serving time. On MetaTool (199~tools, 4,287~queries), this improves NDCG@5 from 0.869 to 0.940; on ToolBench (2,413~APIs), from 0.834 to 0.848. We also evaluate two learned extensions: a 2,625-parameter MLP re-ranker and a 197K-parameter contrastive adapter. The MLP re-ranker hurts or matches baseline when outcome data is sparse relative to the tool set; the contrastive adapter provides comparable gains on MetaTool (NDCG@5: 0.931). All methods are evaluated on the same held-out 30\% test split. The practical takeaway is to start with the zero-cost refinement and add learned components only when data density warrants it. All mechanisms run within single-digit millisecond CPU budgets.

Full-waveform inversion (FWI) is a high-resolution seismic imaging method that estimates subsurface velocity by matching simulated and recorded waveforms. However, FWI is highly nonlinear, prone to cycle skipping, and sensitive to noise, particularly when low frequencies are missing or the initial model is poor, leading to failures under imperfect acquisition. Diffusion-regularized FWI introduces generative priors to encourage geologically realistic models, but these priors typically require costly offline pretraining and can deteriorate under distribution shift. Moreover, they assume Gaussian initialization and a fixed noise schedule, in which it is unclear how to map a deterministic FWI iterate and its starting model to a well-defined diffusion time or noise level. To address these limitations, we introduce Self-Flow-Matching assisted Full-Waveform Inversion (SFM-FWI), a physics-driven framework that eliminates the need for large-scale offline pretraining while avoiding the noise-level alignment ambiguity. SFM-FWI leverages flow matching to learn a transport field without assuming Gaussian initialization or a predefined noise schedule, so the initial model can be used directly as the starting point of the dynamics. Our approach trains a single flow network online using the governing physics and observed data. At each outer iteration, we build an interpolated model and update the flow by backpropagating the FWI data misfit, providing self-supervision without external training pairs. Experiments on challenging synthetic benchmarks show that SFM-FWI delivers more accurate reconstructions, greater noise robustness, and more stable convergence than standard FWI and pretraining-free regularization methods.

Prompt injection remains one of the most practical attack vectors against LLM-integrated applications. We replicate the Microsoft LLMail-Inject benchmark (Greshake et al., 2024) against current generation models running inside OpenClaw, an open source multitool agent platform. Our proposed defense combines two mechanisms: agent isolation, implemented as a privilege separated two-agent pipeline with tool partitioning, and JSON formatting, which produces structured output that strips persuasive framing before the action agent processes it. We run four experiments on the same 649 attacks that succeeded against our single-agent baseline. The full pipeline achieves 0 percent attack success rate (ASR) on the evaluated benchmark. Agent isolation alone achieves 0.31 percent ASR, approximately 323 times lower than the baseline. JSON formatting alone achieves 14.18 percent ASR, about 7.1 times lower. Our ablation study confirms that agent isolation is the dominant mechanism. JSON formatting provides additional hardening but is not sufficient on its own. The defense is structural: the action agent never receives raw injection content regardless of model behavior on any individual input.

Backpropagation dominates modern machine learning, yet it is not the only principled method for optimizing dynamical systems. We propose Kalman World Models (KWM), a class of learned state-space models trained via recursive Bayesian filtering rather than reverse-mode automatic differentiation. Instead of gradient descent updates, we replace parameter learning with Kalman-style gain adaptation. Training becomes online filtering; error signals become innovations. We further extend this framework to transformer-based large language models (LLMs), where internal activations are treated as latent dynamical states corrected via innovation terms. This yields a gradient-free training and adaptation paradigm grounded in control theory. We derive stability conditions, analyze computational complexity, and provide empirical results on sequence modeling tasks demonstrating competitive performance with improved robustness and continual adaptation properties.

Low-dose CT (LDCT) reduces radiation exposure but introduces protocol-dependent noise and artifacts that vary across institutions. While federated learning enables collaborative training without centralizing patient data, existing methods personalize in image space, making it difficult to separate scanner noise from patient anatomy. We propose ProFed (Projection Guided Personalized Federated Learning), a framework that complements the image space approach by performing dual-level personalization in the projection space, where noise originates during CT measurements before reconstruction combines protocol and anatomy effects. ProFed introduces: (i) anatomy-aware and protocol-aware networks that personalize CT reconstruction to patient and scanner-specific features, (ii) multi-constraint projection losses that enforce consistency with CT measurements, and (iii) uncertainty-guided selective aggregation that weights clients by prediction confidence. Extensive experiments on the Mayo Clinic 2016 dataset demonstrate that ProFed achieves 42.56 dB PSNR with CNN backbones and 44.83 dB with Transformers, outperforming 11 federated learning baselines, including the physics-informed SCAN-PhysFed by +1.42 dB.

Generative models based on the Flow Matching objective, particularly Rectified Flow, have emerged as a dominant paradigm for efficient, high-fidelity image synthesis. However, while existing research heavily prioritizes generation quality and architectural scaling, the underlying dynamics of how RF models memorize training data remain largely underexplored. In this paper, we systematically investigate the memorization behaviors of RF through the test statistics of Membership Inference Attacks (MIA). We progressively formulate three test statistics, culminating in a complexity-calibrated metric ($T_\text{mc\_cal}$) that successfully decouples intrinsic image spatial complexity from genuine memorization signals. This calibration yields a significant performance surge -- boosting attack AUC by up to 15\% and the privacy-critical TPR@1\%FPR metric by up to 45\% -- establishing the first non-trivial MIA specifically tailored for RF. Leveraging these refined metrics, we uncover a distinct temporal pattern: under standard uniform temporal training, a model's susceptibility to MIA strictly peaks at the integration midpoint, a phenomenon we justify via the network's forced deviation from linear approximations. Finally, we demonstrate that substituting uniform timestep sampling with a Symmetric Exponential (U-shaped) distribution effectively minimizes exposure to vulnerable intermediate timesteps. Extensive evaluations across three datasets confirm that this temporal regularization suppresses memorization while preserving generative fidelity.

Suffix jailbreak attacks serve as a systematic method for red-teaming Large Language Models (LLMs) but suffer from prohibitive computational costs, as a large number of candidate suffixes need to be evaluated before identifying a jailbreak suffix. This paper presents Prefix-Shared KV Cache (PSKV), a plug-and-play inference optimization technique tailored for jailbreak suffix generation. Our method is motivated by a key observation that when performing suffix jailbreaking, while a large number of candidate prompts need to be evaluated, they share the same targeted harmful instruction as the prefix. Therefore, instead of performing redundant inference on the duplicated prefix, PSKV maintains a single KV cache for this prefix and shares it with every candidate prompt, enabling the parallel inference of diverse suffixes with minimal memory overhead. This design enables more aggressive batching strategies that would otherwise be limited by memory constraints. Extensive experiments on six widely used suffix attacks across five widely deployed LLMs demonstrate that PSKV reduces inference time by 40\% and peak memory usage by 50\%, while maintaining the original Attack Success Rate (ASR). The code has been submitted and will be released publicly.

Standard evaluation metrics suggest that Denoising Diffusion Models based on U-Net or Transformer architectures generalize well in practice. However, as it can be shown that an optimal Diffusion Model fully memorizes the training data, the model error determines generalization. Here, we show that although sufficiently large denoiser models show increasing memorization of the training set with increasing training time, the resulting denoising trajectories do not follow this trend. Our experiments indicate that the reason for this observation is rooted in the fact that overfitting occurs at intermediate noise levels, but the distribution of noisy training data at these noise levels has little overlap with denoising trajectories during inference. To gain more insight, we make use of a 2D toy diffusion model to show that overfitting at intermediate noise levels is largely determined by model error and the density of the data support. While the optimal denoising flow field localizes sharply around training samples, sufficient model error or dense support on the data manifold suppresses exact recall, yielding a smooth, generalizing flow field. To further support our results, we investigate how several factors, such as training time, model size, dataset size, condition granularity, and diffusion guidance, influence generalization behavior.

Model stitching, connecting early layers of one model (source) to later layers of another (target) via a light stitch layer, has served as a probe of representational compatibility. Prior work finds that models trained on the same dataset remain stitchable (negligible accuracy drop) despite different initializations or objectives. We revisit stitching for Vision Foundation Models (VFMs) that vary in objectives, data, and modality mix (e.g., CLIP, DINOv2, SigLIP 2) and ask: Are heterogeneous VFMs stitchable? We introduce a systematic protocol spanning the stitch points, stitch layer families, training losses, and downstream tasks. Three findings emerge. (1) Stitch layer training matters: conventional approaches that match the intermediate features at the stitch point or optimize the task loss end-to-end struggle to retain accuracy, especially at shallow stitch points. (2) With a simple feature-matching loss at the target model's penultimate layer, heterogeneous VFMs become reliably stitchable across vision tasks. (3) For deep stitch points, the stitched model can surpass either constituent model at only a small inference overhead (for the stitch layer). Building on these findings, we further propose the VFM Stitch Tree (VST), which shares early layers across VFMs while retaining their later layers, yielding a controllable accuracy-latency trade-off for multimodal LLMs that often leverage multiple VFMs. Taken together, our study elevates stitching from a diagnostic probe to a practical recipe for integrating complementary VFM strengths and pinpointing where their representations align or diverge.

Structured pruning is widely used to compress large language models (LLMs), yet its effectiveness depends heavily on neuron importance estimation. Most existing methods estimate neuron importance from activation statistics on a single calibration dataset, which introduces calibration bias and degrades downstream cross-task generalization. We observe that neurons exhibit heterogeneous distribution sensitivity, with distribution-robust neurons maintaining consistent rankings across datasets and distribution-sensitive neurons showing high cross-dataset ranking variance. Based on this, we identify two structural limitations in existing methods. First, ranking all neurons within a shared space causes distribution-sensitive neurons that strongly activate on calibration inputs to dominate, crowding out distribution-robust neurons critical for out-of-distribution tasks. Second, applying activation-based importance metrics uniformly can be unreliable. Distribution-sensitive neurons that infrequently activate on calibration data receive insufficient activation signal for accurate local ranking. To address these limitations, we propose GPrune-LLM, a generalization-aware structured pruning framework that explicitly accounts for neuron differences in cross-distribution behavior. We first partition neurons into behavior-consistent modules to localize ranking competition, then evaluate activation-based metric reliability per module according to distribution sensitivity and score magnitude. For modules where activation-based scoring is unreliable, we switch to an activation-independent metric. Finally, we adaptively learn module-wise sparsity. Extensive experiments across multiple downstream tasks demonstrate GPrune-LLM's consistent improvements in post-compression generalization, particularly at high sparsity, and reduced dependence on importance metric choice.

Recently, Multimodal Large Language Models (MLLMs) have been widely integrated into diffusion frameworks primarily as text encoders to tackle complex tasks such as spatial reasoning. However, this paradigm suffers from two critical limitations: (i) MLLMs text encoder exhibits insufficient reasoning depth. Single-step encoding fails to activate the Chain-of-Thought process, which is essential for MLLMs to provide accurate guidance for complex tasks. (ii) The guidance remains invariant during the decoding process. Invariant guidance during decoding prevents DiT from progressively decomposing complex instructions into actionable denoising steps, even with correct MLLM encodings. To this end, we propose Endogenous Chain-of-Thought (EndoCoT), a novel framework that first activates MLLMs' reasoning potential by iteratively refining latent thought states through an iterative thought guidance module, and then bridges these states to the DiT's denoising process. Second, a terminal thought grounding module is applied to ensure the reasoning trajectory remains grounded in textual supervision by aligning the final state with ground-truth answers. With these two components, the MLLM text encoder delivers meticulously reasoned guidance, enabling the DiT to execute it progressively and ultimately solve complex tasks in a step-by-step manner. Extensive evaluations across diverse benchmarks (e.g., Maze, TSP, VSP, and Sudoku) achieve an average accuracy of 92.1%, outperforming the strongest baseline by 8.3 percentage points. The code and dataset are publicly available at https://lennoxdai.github.io/EndoCoT-Webpage/.

Intelligent systems across physics, language and perception often exhibit factorisable structure, yet are typically modelled by monolithic neural architectures that do not explicitly exploit this structure. The separable neural architecture (SNA) addresses this by formalising a representational class that unifies additive, quadratic and tensor-decomposed neural models. By constraining interaction order and tensor rank, SNAs impose a structural inductive bias that factorises high-dimensional mappings into low-arity components. Separability need not be a property of the system itself: it often emerges in the coordinates or representations through which the system is expressed. Crucially, this coordinate-aware formulation reveals a structural analogy between chaotic spatiotemporal dynamics and linguistic autoregression. By treating continuous physical states as smooth, separable embeddings, SNAs enable distributional modelling of chaotic systems. This approach mitigates the nonphysical drift characteristics of deterministic operators whilst remaining applicable to discrete sequences. The compositional versatility of this approach is demonstrated across four domains: autonomous waypoint navigation via reinforcement learning, inverse generation of multifunctional microstructures, distributional modelling of turbulent flow and neural language modelling. These results establish the separable neural architecture as a domain-agnostic primitive for predictive and generative intelligence, capable of unifying both deterministic and distributional representations.

The Model Context Protocol (MCP) standardizes how AI agents discover and invoke external tools, with over 10,000 active servers and 97 million monthly SDK downloads as of early 2026. Yet MCP does not yet standardize how agents safely operate those tools at production scale. Three protocol-level primitives remain missing: identity propagation, adaptive tool budgeting, and structured error semantics. This paper identifies these gaps through field lessons from an enterprise deployment of an AI agent platform integrated with a major cloud provider's MCP servers (client name redacted). We propose three mechanisms to fill them: (1) the Context-Aware Broker Protocol (CABP), which extends JSON-RPC with identity-scoped request routing via a six-stage broker pipeline; (2) Adaptive Timeout Budget Allocation (ATBA), which frames sequential tool invocation as a budget allocation problem over heterogeneous latency distributions; and (3) the Structured Error Recovery Framework (SERF), which provides machine-readable failure semantics that enable deterministic agent self-correction. We organize production failure modes into five design dimensions (server contracts, user context, timeouts, errors, and observability), document concrete failure vignettes, and present a production readiness checklist. All three algorithms are formalized as testable hypotheses with reproducible experimental methodology. Field observations demonstrate that while MCP provides a solid protocol foundation, reliable agent tool integration requires infrastructure-level mechanisms that the specification does not yet address.

Synthetic data has become essential for training code generation models, yet it introduces significant noise and hallucinations that are difficult to detect with current metrics. Existing data selection methods like Instruction-Following Difficulty (IFD) typically assess how hard a model generates an answer given a query ($A|Q$). However, this metric is ambiguous on noisy synthetic data, where low probability can distinguish between intrinsic task complexity and model-generated hallucinations. Here, we propose QAQ, a novel data selection framework that evaluates data quality from the reverse direction: how well can the answer predict the query ($Q|A$)? We define Reverse Mutual Information (RMI) to quantify the information gain about the query conditioned on the answer. Our analyses reveal that both extremes of RMI signal quality issues: low RMI indicates semantic misalignment, while excessively high RMI may contain defect patterns that LLMs easily recognize. Furthermore, we introduce a selection strategy based on the disagreement between strong and weak models to identify samples that are valid yet challenging. Experiments on the WarriorCoder dataset demonstrate that selecting just 25% of data using stratified RMI achieves comparable performance to full-data training, significantly outperforming existing data selection methods. Our approach highlights the importance of bidirectional semantic coherence in synthetic data curation, offering a scalable pathway to reduce computational costs without sacrificing model capability.

Longitudinal brain MRI is essential for characterizing the progression of neurological diseases such as Alzheimer's disease assessment. However, current deep-learning tools fragment this process: classifiers reduce a scan to a label, volumetric pipelines produce uninterpreted measurements, and vision-language models (VLMs) may generate fluent but potentially hallucinated conclusions. We present LoV3D, a pipeline for training 3D vision-language models, which reads longitudinal T1-weighted brain MRI, produces a region-level anatomical assessment, conducts longitudinal comparison with the prior scan, and finally outputs a three-class diagnosis (Cognitively Normal, Mild Cognitive Impairment, or Dementia) along with a synthesized diagnostic summary. The stepped pipeline grounds the final diagnosis by enforcing label consistency, longitudinal coherence, and biological plausibility, thereby reducing the risks of hallucinations. The training process introduces a clinically-weighted Verifier that scores candidate outputs automatically against normative references derived from standardized volume metrics, driving Direct Preference Optimization without a single human annotation. On a subject-level held-out ADNI test set (479 scans, 258 subjects), LoV3D achieves 93.7% three-class diagnostic accuracy (+34.8% over the no-grounding baseline), 97.2% on two-class diagnosis accuracy (+4% over the SOTA) and 82.6% region-level anatomical classification accuracy (+33.1% over VLM baselines). Zero-shot transfer yields 95.4% on MIRIAD (100% Dementia recall) and 82.9% three-class accuracy on AIBL, confirming high generalizability across sites, scanners, and populations. Code is available at https://github.com/Anonymous-TEVC/LoV-3D.

Formal verification of memory-manipulating programs critically depends on precise function specifications that capture memory states written by experts. This requirement has become a major bottleneck as large language models (LLMs) increasingly generate low-level systems code whose correctness cannot be assumed. To enable scalable formal verification, we focus exclusively on function specification generation, deliberately avoiding the synthesis of complex loop invariants that are central to traditional verification pipelines. We propose a neuro-symbolic framework for automatically generating memory-aware formal function specifications for C programs from natural language problem descriptions and function signatures. The pipeline first produces candidate specifications via in-context learning, and then iteratively refines them using compiler diagnostics from symbolic provers and the verification toolchain. In particular, we validate candidate specifications by constructing a proof for the negation of the specification with concrete examples, enabling machine-checked rejection of plausible-but-incorrect specifications. To support systematic evaluation, we introduce LeetCode-C-Spec, a new benchmark of 200 C programming problems for generating memory-aware formal function specifications. Experiments show that iterative refinement substantially improves syntactic validity, while symbolic prover-based refutation significantly enhances correctness assessment by filtering false positives that LLM-only judges frequently accept. Our results demonstrate that combining neural generation with symbolic feedback provides an effective approach to formal specification synthesis for memory-safe systems software.

Technology organizations continuously invest in professional development, but face difficulties in transferring learning to project practice. This exploratory qualitative study investigates which improvements software engineering professionals suggest for organizational learning processes. 174 open-ended responses were analyzed through reflexive thematic analysis. Five themes emerged: practical applicability and alignment with needs; pedagogical quality and organization; time and structural conditions; incentives and institutional recognition; and interaction, mentoring, and social exchange. The results indicate that improving learning requires systemic interventions that integrate practical relevance, structural support, and a favorable institutional culture.

Large language models (LLMs) have demonstrated remarkable capabilities in code generation, yet their potential for generating kernels specifically for mobile devices remains largely unexplored. In this work, we extend the scope of automated kernel generation to the mobile domain to investigate the central question: Can LLMs write efficient kernels for mobile devices? To enable systematic investigation, we introduce MobileKernelBench, a comprehensive evaluation framework comprising a benchmark prioritizing operator diversity and cross-framework interoperability, coupled with an automated pipeline that bridges the host-device gap for on-device verification. Leveraging this framework, we conduct extensive evaluation on the CPU backend of Mobile Neural Network (MNN), revealing that current LLMs struggle with the engineering complexity and data scarcity inherent to mobile frameworks; standard models and even fine-tuned variants exhibit high compilation failure rates (over 54%) and negligible performance gains due to hallucinations and a lack of domain-specific grounding. To overcome these limitations, we propose the Mobile Kernel Agent (MoKA), a multi-agent system equipped with repository-aware reasoning and a plan-and-execute paradigm. Validated on MobileKernelBench, MoKA achieves state-of-the-art performance, boosting compilation success to 93.7% and enabling 27.4% of generated kernels to deliver measurable speedups over native libraries.

Fuzz testing is one of the most effective techniques for detecting bugs and vulnerabilities in software. However, as the basis of fuzz testing, automated heuristics often fail to uncover deep or complex vulnerabilities. As a result, the performance of fuzz testing remains limited. One promising way to address this limitation is to integrate human expert guidance into the paradigm of fuzz testing. Even though some works have been proposed in this direction, there is still a lack of a systematic research roadmap for combining Human-in-the-Loop (HITL) and fuzz testing, hindering the potential for further enhancing fuzzing effectiveness. To bridge this gap, this paper outlines a forward-looking research roadmap for HITL for fuzz testing. Specifically, we highlight the promise of visualization techniques for interpretable fuzzing processes, as well as on-the-fly interventions that enable experts to guide fuzzing toward hard-to-reach program behaviors. Moreover, the rise of Large Language Models (LLMs) introduces new opportunities and challenges, raising questions about how humans can efficiently provide actionable knowledge, how expert meta-knowledge can be leveraged, and what roles humans should play in the intelligent fuzzing loop with LLMs. To address these questions, we survey existing work on HITL fuzz testing and propose a research agenda emphasizing future opportunities in (1) human monitoring, (2) human steering, and (3) human-LLM collaboration. We call for a paradigm shift toward interactive, human-guided fuzzing systems that integrate expert insight with AI-powered automation in the next-generation fuzzing ecosystem.

Knowledge Graphs (KGs) are widely used to represent structured knowledge, yet their automatic construction, especially with Large Language Models (LLMs), often results in incomplete or noisy outputs. Knowledge Graph Completion (KGC) aims to infer and add missing triples, but most existing methods either rely on structural embeddings that overlook semantics or language models that ignore the graph's structure and depend on external sources. In this work, we present OMNIA, a two-stage approach that bridges structural and semantic reasoning for KGC. It first generates candidate triples by clustering semantically related entities and relations within the KG, then validates them through lightweight embedding filtering followed by LLM-based semantic validation. OMNIA performs on the internal KG, without external sources, and specifically targets implicit semantics that are most frequent in LLM-generated graphs. Extensive experiments on multiple datasets demonstrate that OMNIA significantly improves F1-score compared to traditional embedding-based models. These results highlight OMNIA's effectiveness and efficiency, as its clustering and filtering stages reduce both search space and validation cost while maintaining high-quality completion.

Vision-based fall analysis has advanced rapidly, but a key bottleneck remains: visually similarmotions can correspond to very different physical outcomes because small differences in contactmechanics and protective responses are hard to infer from appearance alone. Most existingapproaches handle this by supervised injury prediction, which depends on reliable injury labels.In practice, such labels are difficult to obtain: video evidence is often ambiguous (occlusion,viewpoint limits), and true injury events are rare and cannot be safely staged, leading to noisysupervision. We address this problem with PHARL (PHysics-aware Alignment RepresentationLearning), which learns physically meaningful fall representations without requiring clinicaloutcome labels. PHARL regularizes motion embeddings with two complementary constraints:(1) trajectory-level temporal consistency for stable representation learning, and (2) multi-classphysics alignment, where simulation-derived contact outcomes shape embedding geometry. Bypairing video windows with temporally aligned simulation descriptors, PHARL captures localimpact-relevant dynamics while keeping inference purely feed-forward. Experiments on fourpublic datasets show that PHARL consistently improves risk-aligned representation quality overvisual-only baselines while maintaining strong fall-detection performance. Notably, PHARL alsoexhibits zero-shot ordinality: an interpretable severity structure (Head > Trunk > Supported)emerges without explicit ordinal supervision.

Emotion recognition in videos is a pivotal task in affective computing, where identifying subtle psychological states such as Ambivalence and Hesitancy holds significant value for behavioral intervention and digital health. Ambivalence and Hesitancy states often manifest through cross-modal inconsistencies such as discrepancies between facial expressions, vocal tones, and textual semantics, posing a substantial challenge for automated recognition. This paper proposes a recognition framework that integrates temporal segment modeling with Multimodal Large Language Models. To address computational efficiency and token constraints in long video processing, we employ a segment-based strategy, partitioning videos into short clips with a maximum duration of 5 seconds. We leverage the Qwen3-Omni-30B-A3B model, fine-tuned on the BAH dataset using LoRA and full-parameter strategies via the MS-Swift framework, enabling the model to synergistically analyze visual and auditory signals. Experimental results demonstrate that the proposed method achieves an accuracy of 85.1% on the test set, significantly outperforming existing benchmarks and validating the superior capability of Multimodal Large Language Models in capturing complex and nuanced emotional conflicts. The code is released at https://github.com/dlnn123/A-H-Detection-with-Qwen-Omni.git.

Semantic correspondence is essential for handling diverse in-the-wild images lacking explicit correspondence annotations. While recent 2D foundation models offer powerful features, adapting them for unsupervised learning via nearest-neighbor pseudo-labels has key limitations: it operates locally, ignoring structural relationships, and consequently its reliance on 2D appearance fails to resolve geometric ambiguities arising from symmetries or repetitive features. In this work, we address this by reformulating pseudo-label generation as a Fused Gromov-Wasserstein (FGW) problem, which jointly optimizes inter-feature similarity and intra-structural consistency. Our framework, Shape-of-You (SoY), leverages a 3D foundation model to define this intra-structure in the geometric space, resolving abovementioned ambiguity. However, since FGW is a computationally prohibitive quadratic problem, we approximate it through anchor-based linearization. The resulting probabilistic transport plan provides a structurally consistent but noisy supervisory signal. Thus, we introduce a soft-target loss dynamically blending guidance from this plan with network predictions to build a learning framework robust to this noise. SoY achieves state-of-the-art performance on SPair-71k and AP-10k datasets, establishing a new benchmark in semantic correspondence without explicit geometric annotations. Code is available at Shape-of-You.

Tool use has become central to modern LLM agents, yet interface design is rarely isolated as an experimental variable. This paper studies whether schema based tool contracts and structured validation diagnostics improve reliability under strict interaction budgets. We evaluate three conditions that preserve identical tool semantics and information content: free form documentation, JSON Schema specifications, and JSON Schema with structured diagnostics. We implement a deterministic software engineering sandbox with logs, metrics, configurations, and repository tasks, and evaluate a fully crossed pilot with one open local model, three seeds, three interface conditions, and four budgets. We report end task success, interface misuse, execution failures, semantic misuse, recovery behavior, and overhead. In this pilot, success remains zero across conditions, while schema conditions reduce interface misuse but not semantic misuse. The evidence supports a precise interpretation that interface formalization improves contract adherence, but semantic action quality and timeout sensitive tasks remain dominant bottlenecks under constrained local inference.

Vision-Language-Action (VLA) systems have shown strong potential for language-driven robotic manipulation. However, scaling them to long-horizon tasks remains challenging. Existing pipelines typically separate data collection, policy learning, and deployment, resulting in heavy reliance on manual environment resets and brittle multi-policy execution. We present RoboClaw, an agentic robotics framework that unifies data collection, policy learning, and task execution under a single VLM-driven controller. At the policy level, RoboClaw introduces Entangled Action Pairs (EAP), which couple forward manipulation behaviors with inverse recovery actions to form self-resetting loops for autonomous data collection. This mechanism enables continuous on-policy data acquisition and iterative policy refinement with minimal human intervention. During deployment, the same agent performs high-level reasoning and dynamically orchestrates learned policy primitives to accomplish long-horizon tasks. By maintaining consistent contextual semantics across collection and execution, RoboClaw reduces mismatch between the two phases and improves multi-policy robustness. Experiments in real-world manipulation tasks demonstrate improved stability and scalability compared to conventional open-loop pipelines, while significantly reducing human effort throughout the robot lifecycle, achieving a 25% improvement in success rate over baseline methods on long-horizon tasks and reducing human time investment by 53.7%.

Graph-based Retrieval-Augmented Generation (GraphRAG) constructs the Knowledge Graph (KG) from external databases to enhance the timeliness and accuracy of Large Language Model (LLM) generations. However, this reliance on external data introduces new attack surfaces. Attackers can inject poisoned texts into databases to manipulate LLMs into producing harmful target responses for attacker-chosen queries. Existing research primarily focuses on attacking conventional RAG systems. However, such methods are ineffective against GraphRAG. This robustness derives from the KG abstraction of GraphRAG, which reorganizes injected text into a graph before retrieval, thereby enabling the LLM to reason based on the restructured context instead of raw poisoned passages. To expose latent security vulnerabilities in GraphRAG, we propose Knowledge Evolution Poison (KEPo), a novel poisoning attack method specifically designed for GraphRAG. For each target query, KEPo first generates a toxic event containing poisoned knowledge based on the target answer. By fabricating event backgrounds and forging knowledge evolution paths from original facts to the toxic event, it then poisons the KG and misleads the LLM into treating the poisoned knowledge as the final result. In multi-target attack scenarios, KEPo further connects multiple attack corpora, enabling their poisoned knowledge to mutually reinforce while expanding the scale of poisoned communities, thereby amplifying attack effectiveness. Experimental results across multiple datasets demonstrate that KEPo achieves state-of-the-art attack success rates for both single-target and multi-target attacks, significantly outperforming previous methods.

Transformers often display an attention sink: probability mass concentrates on a fixed, content-agnostic position. Are sinks a byproduct of the optimization/training regime? Or are they sometimes functionally necessary in softmax Transformers? Are sinks a byproduct of the optimization/training regime? Or are they sometimes functionally necessary in softmax Transformers? We prove that, in some settings, it is the latter: computing a simple trigger-conditional behavior necessarily induces a sink in softmax self-attention models. Our results formalize a familiar intuition: normalization over a probability simplex must force attention to collapse onto a stable anchor to realize a default state (e.g., when the model needs to ignore the input). We instantiate this with a concrete task: when a designated trigger token appears, the model must return the average of all preceding token representations, and otherwise output zero, a task which mirrors the functionality of attention heads in the wild (Barbero et al., 2025; Guo et al., 2024). We also prove that non-normalized ReLU attention can solve the same task without any sink, confirming that the normalization constraint is the fundamental driver of sink behavior. Experiments validate our predictions and demonstrate they extend beyond the theoretically analyzed setting: softmax models develop strong sinks while ReLU attention eliminates them in both single-head and multi-head variants.

Nonlinear Probabilistic Latent Variable Models (NPLVMs) are a cornerstone of soft sensor modeling due to their capacity for uncertainty delineation. However, conventional NPLVMs are trained using amortized variational inference, where neural networks parameterize the variational posterior. While facilitating model implementation, this parameterization converts the distributional optimization problem within an infinite-dimensional function space to parameter optimization within a finite-dimensional parameter space, which introduces an approximation error gap, thereby degrading soft sensor modeling accuracy. To alleviate this issue, we introduce KProxNPLVM, a novel NPLVM that pivots to relaxing the objective itself and improving the NPLVM's performance. Specifically, we first prove the approximation error induced by the conventional approach. Based on this, we design the Wasserstein distance as the proximal operator to relax the learning objective, yielding a new variational inference strategy derived from solving this relaxed optimization problem. Based on this foundation, we provide a rigorous derivation of KProxNPLVM's optimization implementation, prove the convergence of our algorithm can finally sidestep the approximation error, and propose the KProxNPLVM by summarizing the abovementioned content. Finally, extensive experiments on synthetic and real-world industrial datasets are conducted to demonstrate the efficacy of the proposed KProxNPLVM.

We present Verified Multi-Agent Orchestration (VMAO), a framework that coordinates specialized LLM-based agents through a verification-driven iterative loop. Given a complex query, our system decomposes it into a directed acyclic graph (DAG) of sub-questions, executes them through domain-specific agents in parallel, verifies result completeness via LLM-based evaluation, and adaptively replans to address gaps. The key contributions are: (1) dependency-aware parallel execution over a DAG of sub-questions with automatic context propagation, (2) verification-driven adaptive replanning that uses an LLM-based verifier as an orchestration-level coordination signal, and (3) configurable stop conditions that balance answer quality against resource usage. On 25 expert-curated market research queries, VMAO improves answer completeness from 3.1 to 4.2 and source quality from 2.6 to 4.1 (1-5 scale) compared to a single-agent baseline, demonstrating that orchestration-level verification is an effective mechanism for multi-agent quality assurance.

Ramaswamy et al. reported in \textit{Nature Medicine} that ChatGPT Health under-triages 51.6\% of emergencies, concluding that consumer-facing AI triage poses safety risks. However, their evaluation used an exam-style protocol -- forced A/B/C/D output, knowledge suppression, and suppression of clarifying questions -- that differs fundamentally from how consumers use health chatbots. We tested five frontier LLMs (GPT-5.2, Claude Sonnet 4.6, Claude Opus 4.6, Gemini 3 Flash, Gemini 3.1 Pro) on a 17-scenario partial replication bank under constrained (exam-style, 1,275 trials) and naturalistic (patient-style messages, 850 trials) conditions, with targeted ablations and prompt-faithful checks using the authors' released prompts. Naturalistic interaction improved triage accuracy by 6.4 percentage points ($p = 0.015$). Diabetic ketoacidosis was correctly triaged in 100\% of trials across all models and conditions. Asthma triage improved from 48\% to 80\%. The forced A/B/C/D format was the dominant failure mechanism: three models scored 0--24\% with forced choice but 100\% with free text (all $p < 10^{-8}$), consistently recommending emergency care in their own words while the forced-choice format registered under-triage. Prompt-faithful checks on the authors' exact released prompts confirmed the scaffold produces model-dependent, case-dependent results. The headline under-triage rate is highly contingent on evaluation format and should not be interpreted as a stable estimate of deployed triage behavior. Valid evaluation of consumer health AI requires testing under conditions that reflect actual use.

State-of-the-art text-to-video models often look realistic frame-by-frame yet fail on simple interactions: motion starts before contact, actions are not realized, objects drift after placement, and support relations break. We argue this stems from frame-first denoising, which updates latent state everywhere at every step without an explicit notion of when and where an interaction is active. We introduce Event-Driven Video Generation (EVD), a minimal DiT-compatible framework that makes sampling event-grounded: a lightweight event head predicts token-aligned event activity, event-grounded losses couple activity to state change during training, and event-gated sampling (with hysteresis and early-step scheduling) suppresses spurious updates while concentrating updates during interactions. On EVD-Bench, EVD consistently improves human preference and VBench dynamics, substantially reducing failure modes in state persistence, spatial accuracy, support relations, and contact stability without sacrificing appearance. These results indicate that explicit event grounding is a practical abstraction for reducing interaction hallucinations in video generation.

We develop a novel approach to ultrafast optical modulation of quantum-mechanical phenomena at the interface of plasmonic metals. Focusing on efficient and versatile nanoparticle-on-mirror plasmonic nanocavities, we discuss indirect control of plasmonic properties through laser-induced ballistic hot electron injection. Overcoming the limitations precluding the observations of laser-driven mesoscopic phenomena in the time domain with state-of-the-art amplified sources, our proposed experimental approach can be readily realized without irreversible optical damage and holds immense potential for the future development of ultrafast electrodynamics in nanogaps, applications in photochemistry and nanoscale control of quantum emitters. Agreeing with the results of numerical simulations, an intuitive microscopic model for the proposed time-dependent mesoscopic electrodynamics facilitates the analysis of the temperature-induced modulation of quantum plasmonic properties in a broad parameter space. Our work expands the realm of quantum nanophotonics onto non-equilibrium electronic systems and facilitates the development of ultrafast methods in active plasmonics.

Optical thermodynamic theory provides distinct viewpoint to rich set of optical phenomena in multimode optical systems. However, standard theory ignores nonlinear effect, severely limiting its range of application. Within a mean-field approximation, we develop a nonlinear optical thermodynamic theory, taking into account the renormalization of linear spectrum due to inter-mode interaction, reminiscent of the van der Waals theory of gases. The resultant nonlinear equation of state enables prediction of power localization, as well as cooling/heating in optical Joule-Thomson expansion, thus providing a unified thermodynamic perspective on nonlinear control and transport of optical waves.

Scanning Probe Microscopy or SPM offers nanoscale resolution but is frequently marred by structured artefacts such as line scan dropout, gain induced noise, tip convolution, and phase hops. While most available methods treat SPM artefact removal as isolated denoising or interpolation tasks, the generative inpainting perspective remains largely unexplored. In this work, we introduce a diffusion based inpainting framework tailored to scientific grayscale imagery. By fine tuning less than 0.2 percent of BrushNet weights with rank constrained low rank adaptation (LoRA), we adapt a pretrained diffusion model using only 7390 artefact, clean pairs distilled from 739 experimental scans. On our forthcoming public SPM InpBench benchmark, the LoRA enhanced model lifts the Peak Signal to Noise Ratio or PSNR by 6.61 dB and halves the Learned Perceptual Image Patch Similarity or LPIPS relative to zero-shot inference, while matching or slightly surpassing the accuracy of full retraining, trainable on a single GPU instead of four high-memory cards. The approach generalizes across various SPM image channels including height, amplitude and phase, faithfully restores subtle structural details, and suppresses hallucination artefacts inherited from natural image priors. This lightweight framework enables efficient, scalable recovery of irreplaceable SPM images and paves the way for a broader diffusion model adoption in nanoscopic imaging analysis.

We introduce a material-agnostic platform for \emph{universal tuning of quantum electrodynamic interactions from power laws to exponential screening and logarithmic antiscreening}, realized in a dielectric spacer bounded by two gate-tunable two-dimensional conductors. The structured electromagnetic environment is completely specified by the transverse-magnetic and transverse-electric reflection amplitudes \(r_{\mathrm{TM/TE}}(q_\perp,ω)\) of the sheets. Starting from the QED action and a Green-function formulation, we resum the multiple-reflection series and show that the interactions are governed by a discrete set of transverse cavity harmonics. In the transparent limit \(r_{\rm TM}\to 0\), the interactions reduce to bulk power laws \(U(ρ)\propto ρ^{-α}\). In the reflective limit \(|r_{\rm TM}|\to 1\), the \emph{phase/parity} of \(r_{\rm TM}\) selects two qualitatively distinct branches: a Dirichlet/PEC (screening) branch \(r_{\rm TM}\to -1\) that removes the gapless transverse mode and yields an evanescent Bessel-\(K\) function \(U(ρ)\propto e^{-πρ/d}/\sqrt{ρ/d}\) at \(ρ\gg d\), and an opposite Neumann/PMC-like (antiscreening) branch \(r_{\rm TM}\to +1\) that retains a gapless mode and can strongly enhance the long-range tail. Thus, the same heterostructure provides in situ electrical control over both the \emph{range} and the \emph{strength} of mediated interactions.

Accurately simulating extended periodic systems is a central challenge in condensed matter physics. Neural quantum states (NQS) offer expressive wavefunctions for this task but face issues with scalability. In this work, we successfully extend the neural network backflow (NNBF) approach to ab-initio solid-state materials. Building on our scalable optimization framework for molecules [Liu et al., PRB 112, 155162 (2025)], we introduce a two-stage pruning strategy to manage the massive configuration space expansions: by utilizing a computationally cheap, physics-informed importance proxy, we devote exact NNBF amplitude evaluations solely to the most relevant determinants, significantly improving optimization efficiency, energy estimation, and convergence. Our framework achieves state-of-the-art accuracy across diverse solid-state benchmarks. For 1D hydrogen chains, NNBF matches or surpasses DMRG and AFQMC, remains robust in strongly correlated bond-breaking regimes where coupled-cluster methods fail, and smoothly extrapolates to the thermodynamic limit. We further demonstrate its scalability by computing ground-state potential energy curves for 2D graphene and 3D silicon. Finally, ablation studies validate the computational savings of our pruning strategy and highlight the dependence of the NNBF energies on basis sets.

Aqueous solutions of disodium cromoglycate (DSCG), a representative model system for chromonic liquid crystals, exhibit temperature- and concentration-dependent phase behaviors spanning isotropic, nematic, and columnar phases, as well as their coexistence regions. Nastishin et al. (2018) reported that UV irradiation can alter the phase diagram, transforming a nematic phase into a nematic-isotropic biphasic state due to weakened molecular attractions, accompanied by a slow post-irradiation relaxation. Here, we revisit this phenomenon and elucidate the molecular origin of this phase diagram shift: the UV-induced photodegradation of DSCG into specific photodegradation products, which we identify using liquid chromatography-mass spectrometry. Through an integrated approach combining in situ X-ray scattering and polarized optical microscopy, we demonstrate that these degradation products disrupt the self-assembly of DSCG aggregates, thereby expanding the isotropic and biphasic regions in the phase diagram. These findings demonstrate that chromonic assemblies and their phase behaviors are highly sensitive to minor chemical alterations, providing a potential route toward light-controlled self assembly of soft matter.

Recent theoretical and experimental studies revealed spatiotemporal photonic, and acoustic, vortex beams in open space. The spatiotemporal vortex beams carry orbital angular momentum perpendicular to the wave propagation direction. Here, we report spatiotemporal magnonic vortex beams in a confined geometry of a ferromagnetic nanostrip. The spatiotemporal magnonic vortex beam contains immobile phase dislocations and the wave propagates in a zigzag-like route. It is remarkable that the transverse orbital angular momentums, carried by the phase dislocations, are spatially alternating. Our findings are in sharp contrast to the photonic and acoustic counterparts, and open a new area in the study of spatiotemporal vortex beams.

Solid-state nanopore DNA sequencers present mechanical and chemical stability, reusability, and large-scale integrability. However, their development is hindered by the absence of a protein-free mechanism for controlling DNA translocation, which is accomplished by motor proteins in their biological counterparts. Here, we propose and theoretically analyze a protein-independent ratchet mechanism based on the unzipping of double-stranded DNA at the nanopore rim. When the transmembrane bias exceeds a certain threshold, the base pairs mechanically dissociate, allowing one strand to translocate while the other remains upstream. This unzipping process is known to slow DNA motion, suggesting that voltage pulses can trigger individual unzipping events at externally defined times, a concept referred to as digital unzipping. However, the intrinsic unzipping barrier is insufficient to provide the dwell times required for a reliable ionic-current readout; therefore, an additional mechanism is needed to hold the DNA in place between voltage pulses. To overcome this limitation, we introduce a reversible hold mechanism implemented via electrostatic attraction between DNA and a charged nanopore wall, which temporarily immobilizes the strand once the unzipping fork catches on the nanopore rim. Using a statistical-mechanical model, we track the evolution of the mean and variance of DNA position through each ratchet cycle. Analytical expressions for the corresponding error probabilities show that submicrosecond switching of the hold mechanism enables base-by-base stepping with an error rate <5%. These results theoretically demonstrate that digital unzipping combined with a reversible hold mechanism can yield deterministic single-base motion, thus opening a viable route toward all-solid-state nanopore sequencing.

The branching geometry of biological transport networks is characterized by a diameter scaling exponent $α$. Two structural attractors compete: impedance matching ($α\sim 2$) for pulsatile flow and viscous-metabolic minimization ($α= 3$) for steady flow. Neither predicts the empirically observed $α_{\mathrm{exp}} = 2.70 \pm 0.20$ in mammalian arterial trees. Incorporating sub-linear vessel-wall scaling $h(r) \propto r^p$ ($p = 0.77$) into a three-term metabolic cost rigorously breaks Murray's cubic law -- via Cauchy's functional equation -- bounding the static optimum to $α_t \in [2.90, 2.94]$. We formulate a unified network-level Lagrangian balancing wave-reflection penalties against transport-metabolic costs. Because the operational duty cycle $η$ is uncertain over developmental timescales, we cast the optimization as a zero-sum game between network architecture and environment. Von Neumann's minimax theorem -- proved constructively via strict monotonicity of the cost curves -- yields a unique saddle point $(α^*, η^*)$ satisfying an exact equal-cost condition. We further prove $N = 2$ uniquely maximizes the network stiffness ratio $κ_{\mathrm{eff}}(N)$, deriving binary branching as a structural consequence of the framework. For the porcine coronary tree ($G = 11$ generations), $α^* = 2.72$, within $0.1σ$ of morphometric data. Sensitivity analysis confirms $|Δα^*| < 0.01$ across physiological metabolic ranges; the prediction depends critically only on the histological exponent $p$ -- a zero-parameter derivation from fundamental scaling principles.

We investigate how a weak constant force becomes detectable through fluctuations in anomalous transport in strongly heterogeneous media. Rather than focusing on the mean drift, we show that the key signature of the force appears in the variance of the particle displacement. As representative models, we study a biased continuous-time random walk (CTRW) with nearest-neighbor jumps and a biased quenched trap model (QTM) with a power-law waiting-time tail. By analysing the force dependence of the displacement variance, we quantify how fluctuations respond to weak driving. We find that for $α<2$, the response exhibits an observation-time-induced crossover: at fixed bias, the variance initially follows its unbiased scaling and only at later times crosses over to a force-dominated nonequilibrium regime. Equivalently, at fixed observation time $t$, there exists a threshold bias $\varepsilon_c(t)$ separating an apparently equilibrium-like regime from a detectable nonequilibrium response. This threshold decreases with increasing $t$, implying that weaker forces become observable over longer measurement windows. Quenched disorder further lowers the detection threshold compared with CTRW, and the crossover reflects a competition between the finite observation time and the intrinsic relaxation time of the driven heterogeneous system.

We consider the crystallization of a two-dimensional electron system in a perpendicular magnetic field using composite boson theory. There are three possible states to consider: the Hall liquid, the Wigner crystal, and the Hall crystal (a state with both broken translation symmetry and a quantized Hall response). Within composite boson theory, these states map onto a superconductor, a Mott insulator, and a supersolid of composite bosons respectively. We show that when a $ν= 1$ Hall liquid has a sufficiently soft roton, there is a first order transition to a triangular lattice Hall crystal. If we continue to decrease the roton mass, there is a continuous transition from the Hall crystal to a Wigner crystal. {When the Hall crystal exhibits the integer quantum Hall effect,} this transition {is} described by a free Dirac fermion and, at the critical point, the coupling to the phonons of the crystal is irrelevant, {in the {renormalization group} sense}. We extend this analysis to fractional $ν= 1/m$ Hall liquids. There, due to kinetic frustration arising from flux attachment, honeycomb lattice Hall crystals are preferred over triangular ones at intermediate interaction strength.

How stochastic, microscopic events generate deterministic, macroscopic properties is a fundamental question in physics. We address this question by developing a quantum master equation model for concentrated radical solutions, where random molecular collisions govern the magnetic properties of the system. Our theory reveals a simple mechanism: the first-order exchange contribution averages to zero over collisions, while the second-order term survives as an effective ferromagnetic coupling that enhances magnetization. The model captures the experimentally observed trends in magnetic behavior that deviate from conventional theories. Because the mechanism arises from statistical averaging, it may apply to a broader class of soft matter phenomena, including liquid crystals.

Recently, the projection of the electron's spin on its crystal momentum has been proposed as a metric to quantify electronic chirality of Bloch states in crystals, which is expected to affect a wide range of physical properties, such as magnetoelectric and optical responses. However, a direct experimental quantification of this chirality metric over an entire iso-energy surface has remained elusive. Here, we have used spin- and angle-resolved photoemission spectroscopy to directly probe the electronic chirality by measuring the bulk spin texture of Kramers-Weyl and Weyl cones in RhSi, a chiral topological semimetal with strong spin-orbit coupling (SOC). After quantifying the SOC splitting of Weyl cones, we determine their spin direction along different azimuthal angles to extract energy dependent the deviations (up to ~40°) from perfect parallel spin-momentum locking. From these deviations we define an energy-dependent normalized electron chirality density (NECD), a directly accessible metric of bulk electronic chirality. In RhSi, the NECD decreases from 1 at the Kramers-Weyl point to ~0.8 at ~200 meV below it. Finally, we show that this experimentally grounded NECD provides predictive power for magneto-optical and transport responses of chiral materials, exemplified by the longitudinal Edelstein effect.

Suspensions of paramagnetic colloids exhibit kinetic arrest in strong magnetic fields. Through a dissipative process of toggling the field on and off, suspensions self-assemble into dense and dynamic steady-state phases. Based on the domain elongation, alpha- and contour-shapes, and degree of phase separation, we construct a phase diagram using a k-means clustering analysis. We identify six characteristic structural regimes: a structureless phase, an arrested structure, sheets, ribbons, a spiky phase, and a transient fluid-fluid regime. We further report the distribution and alignment of domains and the generality of the results. We model self-assembled domain shapes using an equilibrium mean-field magnetostatic energy calculation, which predicts the surprising emergence of highly-anisotropic structures driven by the sample's confinement.

Using numerical simulations it is shown that a random, athermal pack of soft frictional grains will store an arbitrary waveform that is applied as a small time-dependent shear while the system is slowly compressed. When the system is decompressed at a later time, an approximation of the input waveform is recalled in time-reversed order as shear stresses on the system boundaries. It is shown that this effect depends on friction between the grains, and is independent of some aspects of the friction model. By systematically increasing the complexity of the stored waveform, it is found that a pack of $10^4$ grains can recall any one of 128 different waveforms with 100% classification accuracy and 512 different waveforms with over 90% classification accuracy, as measured by a neural net trained only on the inputs. This type of waveform memory might be observable in other types of athermal random media that form internal contacts when compressed such as crumpled sheets and nest-like fiber assemblies.

Magnetic particle hyperthermia relies on the efficient conversion of magnetic field energy into heat in biomedical applications, yet the microscopic mechanisms governing heat generation within individual particles remain poorly understood. In this study, AC magnetometry experiments are combined with dynamic micromagnetic simulations to connect microstructural features, magnetization dynamics, and macroscopic heat dissipation. Beyond macroscopic heating metrics, the heat generation is resolved at the intra-particle level, uncovering a heterogeneous landscape of localized ''hot spots'' with nanometer spatial and nanosecond temporal resolution. The results demonstrate that grain size acts as a key experimentally tunable parameter, balancing anisotropy disorder and pinning strength, thereby controlling both the magnitude and spatio-temporal distribution of heat release within the particle. In particular, nanoflower architectures composed by larger grains deliver larger heat generation, while the smaller grains offer a deeper intra-particle pinning landscape, which effectively redistributes the heat generation over extended time windows. Together, our results provide a mechanistic framework linking nanoparticle microstructure to magnetic heating and establish design principles for optimizing nanoflowers as magnetic hyperthermia transducers.

We investigate associative-memory behavior in a reconfigurable particle packing programmed by purely local cyclic driving. The system is a two-dimensional bidisperse Lennard--Jones particle assembly with periodic boundaries evolved under athermal quasistatic relaxation. During training, a fixed set of input particles is driven cyclically while output particles are selected on-the-fly by a region-driving rule and driven according to a prescribed flow pattern; during retrieval, only the inputs are driven. Associative-memory performance is quantified by the cosine similarity between realized and target output displacement directions. Unlike physical learning systems with fixed architecture, learning here arises through emergent weight updates: localized rearrangements modify the contact network and reshape the effective mechanical couplings between inputs and outputs. Across task difficulty we identify three regimes. In an easy setting, the intrinsic mechanical response already produces coherent motion in the right-hand region under input-only driving, yielding high performance without training. In a hard setting, the desired mapping conflicts with the dominant collective drift, resulting in low baseline performance and only modest training gains; introducing intermittent relaxation cycles reduces train--retrieval mismatch and improves performance. In an intermediate quadrupolar task, repositioning the input--output geometry stabilizes the desired response and converts initially stochastic trajectories into reproducible learned motions. Together these results identify minimal physical ingredients for association-based functionality in athermally driven particulate media and motivate an association learning phase diagram for reconfigurable matter.

We experimentally investigate how aerotactic bacteria, confined within a thin liquid film between two solid substrates, respond to a controlled oxygen gradient. We find that the total bacterial number density dictates which mechanism dominates the steady-state spatial distribution: wall accumulation or aerotaxis. At low densities, despite receiving oxygen only from one substrate, motile bacteria accumulate at both walls, forming a symmetric distribution. In contrast, pronounced aerotactic migration toward the oxygen-supplying wall emerges as the density increases. Analyzing the temporal evolution of this bacterial distribution reveals that the aerotactic response is driven by a self-generated oxygen gradient induced by collective respiration. Our diffusion-advection model of bacteria and oxygen, accounting for aerotactic migration, hydrodynamic attraction to the walls, and respiration, quantitatively reproduces our experimental observations and provides valuable insights into bacterial self-organization within complex environments.

The motion of a disk spinning to rest after being tipped on its side is a classic example of a finite-time singularity, yet the dominant dissipation mechanism governing this process remains debated. Using stereoscopic high-speed imaging, we study the dynamics of disks with varying mass and radius on different surfaces. We show that the late-time motion near the singularity is governed by viscous air-drag arising from shear in the boundary layer beneath the disk, as evidenced by the mass dependence of the dynamics, measurements in a partial vacuum, and a geometric control using a steel ring. At earlier times, dissipation is dominated by rolling friction, which on glass exhibits an unexpected sublinear scaling with disk mass, suggesting an adhesion-based rolling resistance. These results clarify the dissipation mechanisms underlying the singularity of Euler's disk and have broader implications for rolling-contact systems operating under low loads on smooth surfaces.

The first-return time is the time that it takes a random walker to go back to the initial position for the first time. We study the first-return time when random walkers perform fractional kinetics, specifically fractional diffusion, that is modelled within the framework of the continuous-time random walk on homogeneous space in the uncoupled formulation with Mittag-Leffler distributed waiting-times. We consider both Markovian and non-Markovian settings, as well as any kind of symmetric jump-size distributions, namely with finite or infinite variance. We show that the first-return time density is indeed independent of the jump-size distribution when it is symmetric, and therefore it is affected only by the waiting-time distribution that embodies the memory of the process. We perform our analysis in two cases: first jump then wait and first wait then jump, and we provide several exact results, including the relation between results in the Markovian and non-Markovian settings and the difference between the two cases.

We revisit the transfer-matrix approach to directed polymers in random media and show that a single ensemble of random transfer-matrix products provides a unified realization of the canonical one-point fluctuation laws in $(1+1)$ dimensions. For a fixed disorder realization, the polymer partition function is obtained as a contraction of the same product matrix $W(t)$, and different contractions reproduce the standard KPZ subclasses: Tracy-Widom GUE (point-to-point), GOE (point-to-line), GSE (half-space point-to-point), and Baik-Rains (stationary line-to-point). In each case, we observe $t^{1/3}$ free-energy fluctuation growth and convergence of standardized distributions with low-order cumulants close to the corresponding universal benchmarks. Viewing geometry-dependent subclasses as projections of a single matrix-product ensemble naturally suggests additional observables intrinsic to $W(t)$. As an example, we examine the leading eigenvalue $λ_1(t)$ whose logarithm exhibits $t^{1/3}$ scaling, while its standardized statistics remain distinct from the canonical Tracy-Widom laws within the accessible range. This transfer-matrix perspective thus organizes known KPZ one-point subclasses within a finite-dimensional matrix framework and highlights matrix-level fluctuation observables beyond geometry-selected universality classes.

We investigate the behavior of dynamical systems with nonholonomic constraints when coupled to a thermal bath, focusing on the paradigmatic case of the Chaplygin sleigh. A straightforward Langevin-type approach obtained by naively adding stochastic and dissipative terms to the equations of motion predicts a regime in which useful work can be extracted, violating the second law of thermodynamics. To resolve this paradox, we resort to a physically motivated implementation of the nonholonomic constraint as the limiting case of a viscous interaction. However, at finite temperature, fluctuation-dissipation relations imply that the viscous force has to be complemented with stochastic forces acting at the contact. We show that their incorporation restores compliance with the second law. Therefore, our results place fundamental limits on the physical realizability of idealized nonholonomic constraints.

We propose a novel protocol to realize discrete time-crystal (DTC) order in clean, periodically driven spin-$1/2$ chains. In each drive cycle, a global spin flip is followed by a two-tone flat-band segment. This flat-band segment engineers a fully degenerate Floquet quasienergy spectrum, suppresses thermalization, and stabilizes a robust period-doubled subharmonic response. Using exact time evolution, we identify a pronounced subharmonic peak at half the drive frequency in the Fourier spectrum of the order parameter, thereby providing clear evidence for the emergence of stable DTC. The resulting phase is insensitive to system size, interaction strength, and interaction range; however, it remains sensitive to spin-rotation errors ($\varepsilon_r$), which can destabilize the subharmonic response. Compared with disorder-induced many-body localized (MBL) and disorder-free dynamically many-body localized (DMBL) DTCs, we find that the exact flat-band protocol offers a broader tunability of drive parameters, whereas MBL and DMBL based DTCs are more resistant to $\varepsilon_r$. In particular, the $\varepsilon_r$ sensitivity can be suppressed by incorporating additional spin-spin interactions that have modest deviations from the ideal flat-band protocol. This manifests itself in a robust DTC response over a finite window of spin-coupling strengths and drive frequencies. Our results establish flat-band driving as a versatile and experimentally relevant route to DTC order in disorder-free spin systems and motivate further exploration of non-equilibrium phases.

We extend the work of Roychowdhury (2026) on skewness variations of the logarithmic flux, driven by large GeV flares in FSRQs, to a sample of 18 FSRQs. We find that they can be categorized into three groups, one where the skewness attains a persistent lower value after a large flare, one where it does not, and those where change in skewness is not significant. To provide a theoretical ground for these results, we use the statistical plasmoid model of Fermo et al. (2010) that self-consistently produces large plasmoids through merging which, when gain energy from the reconnection event and are Doppler aligned, produce large flares. We find that a downsampling of our simulation of 1500 runs to 18 statistically reproduces the observed distribution in p-values for change in skewness. We further compute the ensemble Shannon entropy of the system and the skewness, where the entropy is found to decrease at a $3σ$ level in both the groups where skewness either increases or decreases, as a direct evidence of increase in order in the system caused by a flare. We find that the power spectral densities of the simulated light curves are broken-power-laws, resembling a white noise+red noise broken by the typical cooling timescale in our system, in accordance with known blazar variability. We find that our results are robust to a $200-300\%$ change in several fiducial parameters of the simulation. Our stochastic simulation of plasmoids inside a blazar jet self-consistently reproduces key observable statistical properties of blazar GeV light curves.

As semiconductor devices approach the sub-2 nm technology node, identifying the quantum-mechanical limits of contact-resistance scaling becomes imperative; however, the transition from thermionic emission to direct tunneling in this deep nanoscale regime remains experimentally inaccessible and theoretically undefined. Herein, we present a systematic first-principles framework to characterize metal/2D-semiconductor interfaces at the atomic scale and identify their intrinsic contact resistance and tunneling limits. Based on large-scale multi-space density functional theory calculations, we perform ab initio transmission line model (TLM) analyses for monolayer MoS2 contacted by Sc, Ag, Au, and Pd electrodes in both top-contact and edge-contact geometries. This computational procedure reveals a universal transition in resistance scaling from metal-induced gap states-mediated direct tunneling in the sub-10 nm regime to thermionic emission at longer channel lengths. The resulting transition length provides a rigorous first-principles measure of the critical tunneling length, establishing a physically grounded metric for assessing contact quality and the source-to-drain tunneling limit of 2D ballistic transistors. Using the ab initio TLM method, we further identify optimal contact strategies-top contact with low-work-function metals for n-type operation and edge contact with high-work-function metals for p-type operation. Our study introduces a general computational framework for evaluating and comparing 2D semiconductor contacts and offers practical guidelines for engineering low-resistance, scalable contact technologies for next-generation 2D transistors.

Understanding the dynamics of quantum correlations in many-body systems is a central problem in non-equilibrium quantum physics. We study the spread of mixed-state entanglement in a minimal model of quantum chaos, the kicked field Ising model. By combining the replica trick with the space-time duality of the model, we determine the exact spectrum of the partially transposed reduced density matrix. The resulting flat spectrum leads to exact relations between entanglement negativity, odd entropy and Rényi mutual information at early times. Numerical results further demonstrate that for equal tri-partitions and at late times, all entanglement measures saturate to the Haar-random values. In contrast, for unequal tri-partitions Rényi mutual information and negativity vanish at late times, implying that the corresponding reduced density matrix is factorizable. Extensive numerical simulations also show that the relation remains quantitatively valid for generic initial states, leading us to conjecture it for all initial states and all times.

Capillary waves are a classical free-surface phenomenon in fluid mechanics, yet their behavior in chiral fluids remains largely unexplored. We show that odd viscosity breaks the reciprocity of capillary waves. Using linear theory together with fully nonlinear direct numerical simulations, we find that surface tension creates two inequivalent branches of odd capillary waves: a dispersive branch and a quasi-acoustic branch absent in the capillarity-free limit. Their unequal propagation and attenuation transform standing waves into traveling waves and produce an anomalously deep vortical boundary layer. Above a threshold odd viscosity, nonlinear accumulation of vorticity near the surface reverses the induced shear current and drives bulk particles opposite to the wave motion, giving rise to an anti-Stokes drift with no counterpart in conventional fluids. Our results show how combining capillarity with broken parity can be used to control wave propagation and transport at fluid interfaces, opening a route toward one-way fluidic waveguiding and chirality-programmed interfacial flows.

The topological Hall effect, driven by the exchange interaction between conduction electrons and topological magnetic textures such as skyrmions, is a powerful probe for investigating the topological properties of magnetic materials. Typically, this phenomenon arises in systems with broken global inversion symmetry, where Dzyaloshinskii-Moriya interactions stabilize such textures. Here, we report the discovery of an emergent giant topological Hall effect in the twisted Fe3GeTe2 metallic system, which notably preserves the general global inversion symmetry. This effect manifests exclusively within a narrow window of "magic" twist angles ranging from 0.45° to 0.75°, while it is absent identically outside of that range, highlighting its unique and emergent nature. Micromagnetic simulations reveal that this topological Hall effect originates from a skyrmion lattice induced by alternating in-plane and layer-contrasting Dzyaloshinskii-Moriya interactions that result from local inversion symmetry breaking. Our findings underscore twisted Fe3GeTe2 as a versatile platform for engineering and controlling topological magnetic textures in metallic twisted van der Waals magnets, thereby opening up new avenues for next-generation spintronic devices.

We investigate how sign-indefinite quadratic invariants shape turbulent cascades in incompressible flows with broken time-reversal symmetry, where the dynamics supports strongly anisotropic dispersive waves. Focusing on rotating Euler flow and odd-viscous Euler flow, we isolate the wave component study the corresponding weak-turbulence kinetic equation. We show that helicity conservation substantially simplifies the kinetic equation. Fixing the energy flux by a natural gauge choice, we identify the turbulent spectrum as the unique scale-invariant solution that sustains a constant flux of energy from large to small scales. Under a mild approximation motivated by the accumulation of energy near slow modes, we compute the leading angular dependence and uncover an integrable singularity along the slow-mode curve, that agrees with previous results. We then demonstrate that helicity reorganizes cascade directions at the level of resonant triads. Although helicity is globally sign-indefinite, the helical decomposition splits it into sign-definite contributions on each polarization branch. Triads whose three legs lie on the same branch behave as if constrained by a sign-definite invariant and drive an upscale transfer of energy, producing systematic backscatter even when the net cascade is direct. In the helicity-definite limit (single-branch dynamics), the kinetic equation admits an additional scale-invariant solution associated with helicity transport. Finally, we validate the analytical predictions by numerically evaluating the collision integral in the strongly anisotropic limit, revealing a family of stationary solutions in that regime.

The extreme anisotropy of hyperbolic materials enables extreme wave confinement, but it is also associated with an inherent misalignment between phase and energy flow, which complicates device modeling and design. Here we introduce a Minkowski-space approach to describe hyperbolic wave propagation, showing that this complexity is geometric rather than physical. By embedding anisotropy into an effective Lorentzian metric, we establish a rational design framework for hyperbolic interfaces and lenses, and analytically derive their transfer function and resolution limits, enabling ultra-large numerical apertures and deep sub-diffraction focusing. We validate our theory with the design and full-wave modeling of a planar van der Waals polaritonic lens operating in the mid-infrared frequency range.

We investigate the coarsening kinetics in a long-range variant of the Persistent Voter Model in space dimension $d=1$ and 2. In this model agents can hold two confidence levels, normal and zealot. If normal, agents take the opinion of others chosen at distance $r$ with probability $P(r) \propto r^{-α}$, with $α>d$. While in the zealot state, agents keep their own opinion. Normal (zealot) agents can become zealots (normal) if their opinion is equal (different) to that of the chosen neighbour. Through numerical simulations we show that, for any values of $α$, the model belongs to the same universality class of the long-range Ising model quenched to a small (non-zero) temperature, similarly to what was already known for the nearest-neighbor case. For the one-dimensional case, we further develop an analytical treatment, which reproduces the $α$-dependence of the correlation length and the functional form of the correlation function. These results not only confirm that the introduction of opinion inertia mitigates the strong interfacial noise present in the voter model, thus reinstating the basic kinetic mechanism of the Ising model, but also expand the applicability of this correspondence.

Linear response theory lies at the foundation of transport phenomena, predicting that physical systems respond proportionally to weak external forces. Here we show that this principle can break down in a minimal nonequilibrium setting due to state-dependent stochastic resetting. We consider a driven Langevin particle subject to a resetting mechanism whose rate grows as a power of the particle velocity, motivated by transport processes where faster carriers experience more frequent scattering events. We derive the exact steady-state velocity distribution and establish a moment balance relation that links external driving, viscous dissipation, and resetting-induced dissipation. This relation reveals that the response is controlled by a nonlinear coupling between the velocity and the resetting rate, leading to nonlinear transport. In particular, the mean velocity obeys the exact power law $\langle v\rangle \propto F^{1/(α+1)}$, where $α$ characterizes the velocity dependence of the resetting rate. Our results provide a solvable example in which linear response fails at the level of the leading-order behavior and identify velocity-dependent resetting as a minimal dynamical mechanism for generating nonlinear transport in nonequilibrium steady states.

We prove that free fermions in the spin (tensor product) factorization violate monogamy of mutual information: $I_3^{\mathrm{spin}} > 0$ for three adjacent strips of width $w = 1, 2$ at all Fermi momenta, and for all~$w$ at $z = k_F w < z^* \approx 1.329$. The proof rests on an exact operator identity -- the fermionic and spin reduced density matrices of disjoint regions differ by the parity insertion $(-1)^{N_B}$ in the partial trace -- and a rigorous entropy bound. DMRG calculations on the $t$-$V$ chain quantify the effect for interacting fermions: the factorization contribution to the apparent $K$-dependence of $I_3$ exceeds the genuine interaction contribution by a factor of~8 at moderate filling, and accounts for ${\sim}80\%$ of the deviation observed in spin-basis numerics. Strong repulsion ($K \lesssim 0.7$) restores monogamy in both algebras. These results imply that any use of $I_3$ as a diagnostic -- whether for holographic duality, quantum chaos, or Fermi surface topology -- must specify the operator algebra; without this specification, the sign of $I_3$ is ambiguous.

We study the protection of information in nearly critical topological quantum codes, constructed by perturbing topological stabilizer codes towards continuous quantum phase transitions. Our focus is on the transverse-field toric code subjected to local Pauli decoherence. Despite the strong quantum fluctuations of anyons when the transverse field is tuned infinitesimally close to the critical point, we show that a finite strength of Pauli decoherence remains necessary to irreversibly destroy information encoded in the ground-state manifold. Using a replica statistical physics mapping for the coherent information, we show that decoherence can be understood as introducing a two-dimensional inter-replica defect within a three-dimensional replica statistical physics model. A field theoretical analysis shows that this defect is perturbatively irrelevant to the bulk critical point, and cannot renormalize the transverse field strength, leading to a finite error threshold. We argue that a qualitatively similar conclusion can be drawn for a broad class of nearly critical topological codes, under a variety of decoherence channels.

This work combines theory and experiment to examine the mechanisms underlying the harmonic generation of magnons. We develop a nonlinear spin-wave framework that is directly analogous to harmonic generation in nonlinear optics, and combine it with scanning nitrogen-vacancy (NV) center magnetometry to image and quantify magnonic harmonic generation in a Ni$_{81}$Fe$_{19}$/Pt microstripe. Within this framework, the harmonic response arises from nonlinear magnetization dynamics localized at strongly inhomogeneous textures, such as the sample edges and domain walls, that act as anharmonic confining potentials. Scanning probe imaging confirms that the harmonic response is correspondingly nonuniform and concentrated near the sample edges. We measure an expected nonlinear power-law scaling, a systematic shift toward larger wavevector excitations at higher harmonic order, and a spin-selective response indicative of an increasingly chiral harmonic stray field. These results provide a microscopic understanding of magnonic harmonic generation and highlight its potential for engineering nonlinear functionality in magnonic systems.

Functional properties of ferroelectric thin films are governed by domains that can be engineered by epitaxial strain. Soft X-ray microscopy can image domain structures with elemental and electronic sensitivity, but hitherto its application to strain-stabilized domains has been hindered by the absorption of soft X-rays in epitaxial substrates. Here, it is demonstrated how this limitation can be overcome by locally back-thinning the (110) TbScO$_3$ substrate of epitaxial K$_{0.6}$Na$_{0.4}$NbO$_3$ ferroelectric thin films to achieve soft X-ray transparency at the O K-edge around 530 eV. Strain-induced ferroelectric stripe domains with periods down to 44 nm were resolved by scanning transmission X-ray microscopy and coherent diffractive imaging by exploiting the X-ray linear dichroism of hybridized O 2p-Nb 4d states, providing sensitivity to in-plane polarization components under normal incidence. The results establish soft X-ray microscopy for nanoscale imaging of epitaxial ferroelectric domains structures and open perspectives for time-resolved studies thereof.

Topological Quantum Computing has largely evolved towards a paradigm of manipulating edge localized Majorana within $p$-wave topological superconducting nanowires. To bridge the gap between physical qubit systems and quantum algorithms, we perform a dynamical analysis to extend what is known in the adiabatic regime, providing time-dependent gate elements for further qubit and algorithm modeling efforts. Our analysis covers dynamical considerations for two methods of shuttling domain edge bound Majoranas in a single nanowire system which both function by applying spatiotemporally dependent onsite and hopping parameters within the system's Hamiltonian. We then complicate this model by converting it into the T-qubit to calculate the finite-time gate representation of the shuttling techniques used in a more practical setting. These contributions provide insight for realistic experimental setups in the next-generation of qubit implementation and will hopefully facilitate fault tolerant scalable systems and universal gate design.

The potential for using boundary functionals of random risk processes to solve nuclear safety problems at nuclear power plants is assessed. In certain situations (MSRs (Molten Salt Reactors), High-Temperature Gas-Cooled Reactors (HTGRs), pulverized fuel reactors, reactor startups, and accident analysis (core collapse)), neutron behavior changes significantly. Neutron clustering begins to play an important role, and the distributions characterizing neutron behavior change. The normal distribution is replaced by stable, but also limiting, distributions. Boundary functionals allow for precise calculation of the power quantile and provide a mathematical bridge between abstract directed percolation and engineering calculations of protection settings.

In their 1972 study of approach to equilibrium, Lanford and Robinson showed that gauge-invariant quasi-free states of lattice fermions maximize entropy among all translation-invariant states with a fixed two-point function, and suggested that the maximizer is unique. In subsequent work on this topic, the uniqueness question re-emerged, together with the problem of whether such quasi-free states are weak Gibbs states. We provide a positive answer to both questions within a class of states whose momentum-space two-point function $\widehat C$ satisfies $0<\widehat C(k)<1$ and belongs to the Wiener algebra of the Brillouin zone. The proof reveals that both the entropy maximization principle and weak Gibbsianity follow directly from the thermodynamic formalism for lattice fermions.

The universal generalising approach for non-isothermal behaviour of different alloys has been provided together with the novel deformation modelling. Strong correlation between the model approach and experimental results is shown that permits estimation of main applied parameters such as the linear thermal expansion coefficient and others. Necking contours and critical thickness at corrugation for ribbon and rod specimens are also calculated. Fractal analysis of corrugation folds (their main size) has been carried out for polycrystalline and amorphous ribbon specimens. Structural peculiarities at the plastic deformation stage are investigated with microscopy.

Numerical simulations based on radial basis functions have been developed for systems with complex geometries and have been successfully applied across various fields, including seismology, coastal hydrodynamics, and biology. However, examples in liquid crystal modeling are limited. In this study, we present a Landau-de Gennes numerical simulation of nematic liquid crystals utilizing radial basis functions, emphasizing its advantages over traditional cubic grid calculations, such as enhanced geometric flexibility and improved computational efficiency. Through simulations of liquid crystal-colloid systems with diverse geometries, we demonstrate that our approach effectively captures the essential topological and energetic features of liquid crystal equilibrium structures. Additionally, we introduce an adaptive node refinement scheme that is crucial for resolving the fine structure of singular defects in nematic liquid crystals.

Channel-mediated transport is ubiquitous in biology. A series of works by different theoreticians have sought to determine how the diffusive flux through a channel depends on (a) stochastic gating, (b) channel geometry, and (c) heterogeneous diffusion. In this paper, we derive an explicit estimate for the diffusive flux through a channel that accounts for these three factors. We show that our estimate is exact in certain parameter regimes. We further use stochastic simulations to confirm that our estimate remains accurate across a very broad range of parameters. Our estimate differs from some results in the physics literature.

In recent years, interferometry experiments in fractional quantum Hall devices have reported signatures of a fractional braiding phase for quasiparticles. It was noted, however, that the braiding phase alone does not uniquely determine the exchange phase because of a $π$-ambiguity. Here we analyze a Hanbury Brown-Twiss interferometer in a cross geometry that provides direct access to the fractional exchange phase. Using a non-equilibrium Keldysh calculation in an experimentally relevant regime, we show that the exchange phase can be obtained as the phase shift between Aharonov-Bohm oscillations in a single-particle interference current and those in the current cross-correlation arising from two-particle interference.

We study information flow on a weighted graph whose topology evolves according to a spectral curvature measure $\mathcal{R}$. The model (FIU) defines $\mathcal{R}$ from the diagonal of the graph Green function, propagates energy with curvature-dependent dissipation, and creates long-range links between high-$\mathcal{R}$ nodes at a rate controlled by a coupling parameter $g$. We report three results. First, the system exhibits a sharp phase transition at $g_c \approx 0.023$: below $g_c$, local information flux $σ$ and structure formation are anti-correlated; above $g_c$, they become strongly correlated (Pearson $r \approx 0.75$, $p < 10^{-38}$), with signatures of a continuous transition and mean-field exponent $ν\approx 0.54$. Second, we identify a node-level discrete Poisson relation $\nabla^2\mathcal{R}(i) = κ\,σ_{\rm prev}(i)$, where $κ$ is stable across parameters (CV $= 3.1\%$ across independent configurations). Mediator analysis reveals this correlation is almost entirely mediated by $\mathcal{R}$ itself, identifying it as the central self-organizing variable. Third, the Poisson relation exhibits spontaneous dimensional sensitivity: in 2D lattices both signals decay for $N \gtrsim 576$, while in 3D they persist to $N \lesssim 1728$. This emerges without any dimensional parameter in the rules. The collapse mechanism is curvature homogenization at large $N$. We interpret this as topological frustration in a mesoscopic regime, and discuss analogies with dimensional signatures of gravity.

In this study, we demonstrate the generalizability of graph neural networks in predicting the dynamic heterogeneity of model glass-forming liquids across different temperatures. While previous approaches have often been limited to making predictions at the specific temperatures used during training, we find that our proposed framework - T-BOTAN - enables interpolation to temperatures not included in the training set. We show that the dynamical behavior, the associated four-point correlations, and even the macroscopic temperature can be estimated with sufficient accuracy solely from static particle configurations at untrained temperatures. These results suggest that static configurations encode not only local structural features driving dynamic heterogeneity but also fundamental thermodynamic information.

Valleytronics harnesses the valley degree of freedom -- energy-degenerate extrema in the electronic band structure -- for information storage and processing. Valley Hall effect (VHE) is a cornerstone of valleytronics, enabling electric generation of pure valley currents. While extensively studied in systems with valleys located at time-reversal-breaking points, here, we shift the paradigm to valleytronic platforms with time-reversal-invariant valleys (TRIVs), revealing a novel phenomenon: eccentricity VHE. Unlike conventional VHE, the valley Hall angle for eccentricity VHE is an intrinsic geometric property, governed solely by the eccentricity of the valley Fermi surface, rendering it highly robust against variations in temperature or carrier density. Eccentricity VHE emerges universally across all 25 layer groups supporting TRIVs. We demonstrate these distinctive features in monolayer GeS$_{2}$ via first-principles calculations, predicting a significant valley Hall angle of 0.74. This effect can be detected through nonlocal transport measurements exhibiting characteristic scaling behavior, or, in certain cases, through valley-layer coupling. Our findings reveal a critical overlooked facet of valley Hall physics, transcend the established VHE paradigm, and significantly broadens the scope of valleytronics.

Collective optical excitations, such as localized surface plasmons in metallic nanoparticles and Mie resonances in high-index dielectrics, play a central role in nanoscale light--matter interactions. When such optical modes interact with electronic transitions in matter under suitable conditions, they can couple strongly, analogous to two coupled harmonic oscillators, forming hybrid light--matter states. In this work, we probe this coupling in core--shell nanoparticles using fast electrons in electron energy-loss (EEL) and cathodoluminescence (CL) spectroscopy. Owing to their highly localized fields, fast electrons can excite modes inaccessible with light-based spectroscopies, including higher-order nonradiative modes, which offer greater field confinement and potentially stronger coupling. Here, we develop an analytical framework to calculate the EEL and CL probabilities for spherical core--shell nanoparticles under aloof and penetrating electron trajectories. This formalism is applied to two representative systems: an excitonic core with a metallic shell, and a silicon core with an excitonic shell. Our main focus is to examine how the electron beam position and velocity affect our ability to probe this coupling. Depending on the electron beam parameters, we find that the spectral signature of strong coupling remains robust in plasmonic nanospheres. In contrast, it can be significantly suppressed or even completely obscured in dielectric nanospheres. Our developed formalism enables a deeper understanding of the coupling mechanisms in electron--light--matter interactions, thereby accelerating progress in single-nanoparticle-based polaritonic studies.

Recently proposed analog solvers based on dynamical systems, such as Ising machines, are promising platforms for large-scale combinatorial optimization. Yet, given the heuristic nature of the field, there is very limited insight on optimality guarantees of the solvers, as well as how parameter schedules shape dynamics and outcomes. Here, we develop a dynamical mean-field framework to analyze Ising-machine dynamics for finding the ground state energy of the Sherrington-Kirkpatrick(SK) model of spin glasses and identify mechanisms that enable rapid convergence to provenly near-optimal energies. For a fixed target energy density Ec, we show that solutions are typically reached within O(1) matrix vector multiplications, indicating constant time complexity. We further delineate theoretical limitations arising from different parameter-scheduling trajectories and demonstrate a pronounced benefit of temperature-only annealing for the Coherent Ising Machine. Building on these insights, we propose a general framework for designing optimized parameter schedules, thereby improving the practical effectiveness of Ising machines for complex optimization tasks. The superior performance of the dynamical solvers is illustrated by the attainment of the ground state energy of the SK model.

We present a molecular dynamics study of the influence of knot complexity and molecular mass on glass formation upon cooling in knotted ring polymer melts. We find that cooperative motion, rigidity, and glassy dynamics can be tuned over a wide range by knots. By leveraging these knotting constraints, we assess the validity of prevalent models of glass formation, including the string model based on cooperative particle motion, the localization model emphasizing fluctuations in local particle mobility, and the shoving model derived from emergent elastic properties in relation to material stiffness. In line with our previous findings on polymeric and other glass-forming liquids, we demonstrate that all these models of glass formation provide a quantitative description of segmental relaxation as a function of knot complexity, molecular mass, and temperature, despite their apparently distinct conceptual foundations. Our study thus provides additional evidence for an underlying unity among various theoretical frameworks and for the presence of quantitative relations between the characteristic properties emphasized by these models. Furthermore, we discuss dynamic and elastic heterogeneities in relation to fragility and stiffness variations of knotted ring polymer melts, with a focus on how these trends relate to other glass-forming liquids where fragility is tuned over a large range.

Large-scale land enclosure for speculative mega-development constitutes a non-equilibrium spatial process whose velocity, topology, and irreversibility remain poorly quantified. We study the Pantai Indah Kapuk 2 (PIK2) coastal mega-development north of Jakarta, Indonesia, using eight years (2017--2024) of Sentinel-2 land-use/land-cover (LULC) data at 10-meter resolution. The landscape is projected onto a Marxian probability simplex partitioning terrestrial pixels into Commons, Agrarian, and Capital fractions. Fisher-Rao (FR) geodesic distances on this simplex identify a transformation pulse of $0.405$~rad/yr during 2019--2020, coinciding with major construction activity. Absorbing Markov chain analysis yields expected absorption times into the built environment of $46.0$~years for cropland and $38.1$~years for tree cover, with a pooled built-area self-retention rate of $96.4\%$. Percolation analysis reveals that a giant connected component containing $89$--$95\%$ of all built pixels persists at occupation probabilities $p \in [0.096, 0.162]$, far below the random percolation threshold $p_c \approx 0.593$, indicating planned rather than stochastic spatial growth. The box-counting fractal dimension of the urban boundary increases from $d_f = 1.316$ to $1.397$, consistent with increasingly irregular frontier expansion. These results suggest that information-geometric and statistical-mechanical tools can characterize the kinematic and topological signatures of capitalist spatial accumulation with quantitative precision.

We develop a simulation framework for electrostatic and transport modeling of 2D Topological insulator field-effect transistor (2D TIFETs), based solely on first-principles calculations using density functional theory (DFT). We find that careful consideration of basis set and symmetry constraints in DFT calculations is crucial for determining critical electric field ($E_c$), defined as the electric field intensity at which the topological phase transition occurs. Using ballistic Landauer-B$ü$ttiker formula and local potential profile, the drain current-gate bias voltage ($I_D$-$V_G$) characteristics were obtained and switching behavior was studied. A comparison with the $\mathbf{k}\cdot\mathbf{p}$ model reveals the necessity of DFT calculations for investigating realistic edge dispersions. Our approach provides an efficient and rigorous simulation methodology for mesoscopic transport in 2D TIFETs.

It is expected that a generic closed many-body system prepared in a well-behaved initial state and subjected to a periodic drive will eventually thermalize, i.e. approach the state of maximal entropy. This property, while compatible with and even demanded by the physical intuition, is much stronger than ergodicity or mixing and is difficult to justify mathematically. We describe an infinite set of classical many-body Floquet systems of algebraic origin for which thermalization of very general initial states can be proved. For example, we show that a Gibbs state of any sufficiently uniform local differentiable Hamiltonian heats up to infinite temperature at long times. We show that in agreement with the physical intuition, the only obstruction to thermalization is the existence of local observables which are periodic in time.

Active particles exhibit self-propulsion, leading to transport behavior that differs fundamentally from passive Brownian motion. In confined or structured domains, activity strongly influence escape probabilities and first-passage behavior. Understanding these effects is essential for describing transport in biological microenvironments, microfluidic devices, and heterogeneous media. In this work, leveraging the backward Fokker--Planck equation, we investigate the splitting probability of active particles in confined domains, focusing on both a one-dimensional interval and a two-dimensional corrugated channel. Analytical solutions are derived for the one-dimensional case in various asymptotic regimes. In corrugated channels with small aspect ratios, we develop a Fick--Jacobs reduction that yields effective transport equations along the axial direction, whereas for finite aspect ratios, the splitting dynamics are characterized numerically. We demonstrate how channel geometry, particle activity, and chirality modulate the likelihood of escape through different boundaries. Our results provide quantitative predictions for the transport of active matter in complex environments and highlight the interplay between confinement and activity.

The dramatic viscosity increase observed in dense suspensions under shear poses a major challenge in our understanding of how microscopic contact mechanics translate into macroscopic flow resistance. Here, we introduce a constraint-counting model that incorporates friction and dimensionality naturally without additional assumptions and allows for collapsing of rheological data onto a universal master curve. In this model, we borrow ideas from dry granular jamming physics and classify contacts as either locked or non-locked to define a single state variable, the constraint ratio, which measures the average strength of mechanical constraint per particle. By identifying the constraint ratio as the key control parameter, our framework provides a unifying route toward predictive modeling and rational design of shear-thickening materials.

We perform an end-to-end benchmark of a hybrid sequential quantum computing (HSQC) solver for higher-order unconstrained binary optimization (HUBO), executed on IBM Heron r3 quantum processors to evaluate the potential of current quantum hardware for combinatorial optimization with sub-second end-to-end runtimes. All reported runtimes include the complete pipeline--from preprocessing to QPU execution and postprocessing--under strict wall-clock accounting. Across 20 benchmark instances, a single hybrid attempt produces high-quality solutions in less than one second, matching the ground-state energy in 14 cases. At the same runtime, CPU-based solvers, including simulated annealing, memetic tabu search, and EasySolve, do not reach the value obtained by HSQC, whereas an enhanced parallel tempering method and the GPU-accelerated solver ABS3 reach or surpass it. These results show that HSQC, executed on a single QPU, can achieve performance competitive with strong classical solvers running on 128 vCPUs or 8 NVIDIA A100 GPUs, while also providing a reproducible system-level benchmark for tracking progress as quantum hardware and hybrid sequential workflows improve.

Colloidal gels, like many other soft and disordered solids derive their mechanical properties not only from the strength of interparticle attraction, but also from the symmetry of the forces that constrain particle motion. While non-central interactions are known to profoundly alter rigidity and elasticity, they are typically introduced through particle anisotropy, surface roughness, or patchy interactions, obscuring their independent role. Here we demonstrate a minimal and geometry-preserving route to emergent non-central forces in colloidal gels by reducing the density of surface-grafted polymer brushes. At low brush density, partial brush interpenetration introduces an effective angular bending rigidity at particle contacts, despite fully isotropic particle geometry. This emergent constraint suppresses local densification, stabilizes low-coordination networks, and produces highly ramified gel structures with enhanced elasticity. Combining experiments, simulations, and mean-field theory, we show that these non-central constraints reorganize structure and mechanics across length scales, shifting gelation boundaries and increasing the elastic modulus by nearly a factor of three. Our results establish surface brush density as a generic control parameter for programming interaction symmetry in soft particulate matter, with implications for rigidity, percolation, and mechanical design in disordered systems.

A key challenge for molecular dynamics simulations is efficient exploration of free energy landscapes over relevant collective variables (CV). Common methods for enhancing sampling become prohibitively inefficient beyond only a few CVs; in the case of the widely-used metadynamics method, the computational cost of evaluating and storing the bias potential grows exponentially with the number of dimensions. Here, we introduce TT-Metadynamics, in which the accumulated sum of Gaussian functions in the original metadynamics method is periodically compressed into a low-rank tensor train (TT) representation. The TT enables efficient memory use and prevents the computational cost of evaluating the bias potential from increasing with simulation time. We present a "sketching" algorithm that allows us to construct the TT with linear scaling in the number of CVs. Applied to benchmark systems with up to 14 CVs, the accuracy of TT-Metadynamics matches or exceeds that of standard metadynamics in long simulations, particularly in systems with high barriers. These results establish TT-Metadynamics as a scalable and effective method for computing free energies that are functions of several CVs.

For decades, Hermiticity was considered an immutable axiom of quantum mechanics, essential for ensuring real energies and unitary evolution. This perspective has shifted radically, driven by the realization that non-Hermitian Hamiltonians provide a powerful effective description of open quantum systems, granting access to unique phenomena such as Exceptional Points and the Non-Hermitian Skin Effect. In this Perspective, we chart the trajectory of this field, moving from its established foundations in single-particle, linear models to the emerging frontier of interacting many-body systems. We first clarify the physical origins of non-Hermitian dynamics, distinguishing between mean-field approximations, conditional "no-click" evolution, and exact Liouvillian dynamics. We then focus on the rich phenomenology arising from the interplay of non-Hermiticity and interactions. We discuss interaction-induced topological phases, the generalization of skin effects to the many-body Hilbert space, and the distinct signatures of dissipative quantum chaos and complexity. Finally, we highlight collective phenomena in nonlinear regimes, including skin solitons and dissipative phase transitions. We also comment on measurement-induced entanglement transitions and their relation to non-Hermitian spectra and topology. By synthesizing these diverse developments, we provide a roadmap for the future of non-Hermitian physics.

Scalable architectures for quantum computing using semiconductor quantum dots require interactions between qubits beyond adjacent quantum dots. Here, we propose using nanowires of electrons to mediate the interaction between two quantum dots. Virtual phonons in the linear chain of electrons can mediate an interaction that gives rise to effective spin-spin coupling of the electrons in distant quantum dots. We find coupling strengths of more than 30 MHz for experimentally realisable parameters in GaAs quantum dots.

Magnons, the quanta of spin waves, have significant potential for use in modern technologies, especially when strongly coupled to another mode for read-out and control. However, while magnons strongly interact with microwave photons via the magnetic-dipole interaction to form hybrid cavity-magnon polariton modes, the weak magnetostrictive magnon-phonon interaction, together with large polariton linewidths dominated by magnon dissipation, has so far restricted magnonic-spheres to the weak-coupling regime. The material-limited magnon dissipation rate in particular has been regarded as an unavoidable limitation in these systems. Here, we surpass this long-standing limitation by implementing an active microwave feedback loop to suppress the linewidth of cavity-magnon polaritons and strongly suppress their effective decay rate below the magnon-limited linewidth, thereby enhancing the polariton-phonon cooperativity from C=1 to C=150. As a key milestone, we achieve normal-mode splitting between a cavity-magnon polariton and a mechanical mode, providing direct evidence of three-mode hybridization among photons, magnons, and phonons. Our results establish feedback as a general route to accessing strong-coupling regimes in systems previously thought to be limited by material properties and hence open new opportunities for coherent control in hybrid quantum systems.

Nonlinear phononics has emerged as a powerful paradigm for the nonthermal control of quantum materials by engineering a conservative potential energy landscape. Here, we show that dissipation can serve as an additional control knob for nonequilibrium states in nonlinear phononics. We reveal a nontrivial role of dissipation by investigating a spin-phonon coupled system driven by circularly polarized light. By tuning the spin relaxation time $τ_s$, the steady state undergoes a transition from a trivial limit cycle to a temporally ordered state, which spontaneously breaks the discrete time-translation symmetry imposed by the drive. In this state, both the spin and phonon angular momentum exhibit persistent oscillations at an emergent frequency $Ω_s$, which is generally incommensurate with the driving frequency. This state is stabilized by a dissipation-induced phase lag between spin and phonon angular momentum that generates a feedback loop sustaining the oscillation. The dissipation-controlled transition can be described within a Landau-type framework using a pseudo-potential, where the order parameter has a $U(1)$ phase symmetry, and its amplitude is proportional to the oscillation amplitude of the phonon angular momentum.

In conventional superconductors, superconductivity is generally suppressed by external magnetic fields due to spin-singlet pairing. Here, we report signatures of in-plane-magnetic-field-induced superconductivity in hexalayer rhombohedral graphene and reveal electric-field control of its depairing behavior. With the application of a small in-plane magnetic field $B_{\parallel}$, a superconducting state emerges within a narrow band along a phase boundary. Its properties evolve continuously with increasing $B_{\parallel}$: the superconducting region progressively shifts toward higher electric field as the $B_{\parallel}$ increases and the transition temperature rises with increasing $B_{\parallel}$. Remarkably, the superconducting state remains robust under $B_{\parallel}$ up to 14 T, far exceeding the conventional Pauli limit. Quantum oscillation measurements further reveal that the superconductivity emerges from nematic Fermi surface reconstruction. These results suggest a spin-polarized superconducting states with unconventional origins.

The ability of modern quantum simulators--both digital and analogue--to generate large ensembles of single-shot projective "snapshots" has opened a data-rich avenue for the study of quantum many-body systems. Unsupervised machine learning analysis of such snapshots has gained traction, with numerous works reconstructing phase diagrams by learning and clustering low-dimensional representations of quantum states. Here, we forgo such representation learning in favour of distance learning: we infer the pairwise distances between quantum states--already sufficient for clustering--directly from snapshots. Specifically, we use a single neural discriminator to estimate Csiszar f-divergences--statistical distances between distributions--in an unsupervised manner. The resulting clusters reveal regimes with different dominant correlations, often coinciding with, but not limited to, conventionally defined phases of matter. Beyond phase-diagram exploration, we connect the infinitesimal limit of the inferred divergences to the Fisher information metric and analyse its finite-size scaling. This yields critical exponents of the discovered transitions and enables snapshot-based analysis of universality classes. We apply distance learning to a diverse set of systems characterised by conventional local order parameters (1D transverse-field and 2D classical Ising models), non-local topological order (extended toric code), and higher-order correlations (fermionic t-J model on a triangular lattice). In all cases, we correctly recover boundaries between distinct correlation regimes and, where applicable, quantitatively match established critical behaviour. Finally, we show that distances to suitably chosen reference snapshot distributions help identify the dominant correlations within the discovered clusters, positioning distance learning as a versatile information-geometric probe of quantum many-body physics.

In this work, we consider partial consumption of food by a forager in presence of a threshold energy level. The forager considered here can survive for $S$ steps without food, namely the survival time. The threshold limits the consumption of food in such a way that, the forager will only consume food, whenever its energy is below the threshold $k$. Due to partial consumption of food, a site containing food may not always be fully depleted, which in turn helps in increasing the lifetime of the forager. It has been observed that, in our case, the lifetime always increases with $k/S$, although there is a transition threshold $k^*$ below which the increase of lifetime is rapid and above is low. The transition threshold $k^* \sim \sqrt{S}$. The lifetime $τ$ shows a power law behavior as $τ\sim S^β$. For $k/S=0$, the value of $β$ is $4/3$, it then jumps above $2$ and decreases gradually to $1.84$ with increasing $k/S$. Other important quantities like number of revisits to a site, food statistics etc. have been studied and these also show some interesting scaling behavior. The collection of sites either fully or partially depleted of food after the death of the forager $N_{eat}$ shows a crossover behaviour for $k/S \sim 0.5$.

The theory of open quantum systems addresses how coupling to external degrees of freedom modifies observables and quantum coherence, a situation central to fundamental condensed-matter research and emerging quantum technologies. Schwinger-Keldysh field theory is a natural framework for both open- and nonequilibrium quantum systems in terms of functional integrals. However, its numerical solution is limited by a cubic scaling with the number of time steps. This is particularly prohibitive for scenarios with widely separated time scales, as is often the case for system and environmental scales. We consider a damped quantum harmonic oscillator as a toy model to study a separation-of-scales ansatz based on Hadamard regularization. A time-stepping algorithm for the Kadanoff-Baym equations on the slow system time-scale is presented that captures both low-temperature non-Markovianity and renormalization effects arising from the much faster environment scale.

Impacts from high-energy particles cause correlated errors in superconducting qubits by increasing the quasiparticle density in the vicinity of the Josephson junctions (JJs). Such errors are particularly harmful as they cannot be easily remedied via conventional error correcting codes. Recent experiments reduced correlated errors by making the difference in superconducting gap energy across the JJ larger than the qubit energy. In this work, we assess gap engineering near the JJ ($δΔ_{\mathrm{JJ}}$) and the capacitor/ground-plane ($δΔ_{\mathrm{M1}}$) by exposing arrays of transmon qubits to two sources of radiation. For $α$-particles from an $^{241}$Am source, we observe $T_1$ errors correlated in space and time, supporting a hypothesis that hadronic cosmic rays are a major contributor to the $10^{-10}$ error floor observed in Ref. 1. For electrons from a pulsed linear accelerator, we observe temporally correlated $T_1$ and $T_2$ errors, this measurement is insensitive to spatial correlations. We observe that the severity of correlated $T_1$ errors is reduced for qubit arrays with a greater degree of gap engineering at the JJ. For both $T_1$ and $T_2$ errors, the recovery time is hastened by an increased $δΔ_{\mathrm{M1}}$, which we attribute to the trapping of quasiparticles into the capacitor/ground-plane. We construct a model of quasiparticle dynamics that qualitatively agrees with our observations. This work reinforces the multifaceted influence of radiation on superconducting qubits and provides strategies for improving radiation resilience.

The empirical speed of travelling reaction-diffusion fronts fluctuates due to the intrinsic shot noise of the reactions and diffusion. Here we study the long-time front speed fluctuations of a stochastic Huxley-Zel'dovich front. It involves a population of particles $A$ which perform a fast continuous-time random walk on a one-dimensional lattice and undergo reversible on-site reactions $2A \rightleftarrows 3A$. This front describes an invasion of $A$-particles into an initially empty region of space which, in a deterministic description, is marginally stable but nonlinearly unstable with a zero instability threshold. Typical fluctuations of this front can be described as front diffusion in a reference frame moving with the average front speed. According to the existing perturbation theory, the shot-noise-induced systematic shift of the average front speed, $δc$, and the front diffusion coefficient, $D_f$, are both expected to scale with $N$ as $1/N$, where $N \gg 1$ is the typical number of particles in the transition region. Furthermore, $D_f$ can be determined perturbatively in the small parameter $1/\sqrt{N}$. Our Monte Carlo simulations support these asymptotic results, but also reveal a long-lived anomalous behavior of the first few particles before they reach the expected diffusion asymptotic. We also study large deviations of the empirical speed of the front at long times. These are dominated by optimal histories of the system in the form of a propagating front which travels with a speed different from the average speed, or even travel in the wrong direction.

Graphene is widely recognized for its ultrafast and broadband photocurrent response, but whether the broadband ultrafast characteristics are preserved at mid-infrared wavelengths with photon energies below the optical phonon energy remains an open question. Here, we investigate the carrier dynamics in graphene junctions under mid-infrared excitation using an ultrafast pump-probe photocurrent spectroscopy. We utilize dual split gate devices to demonstrate that the photo-thermoelectric effect can dominate the photoresponse of graphene also for a mid-infrared femtosecond excitation. We observe that graphene retains its broadband photocurrent response in this spectral region, but the photocurrent relaxation time increases from ca. 2 ps below 8-9 micrometer up to 3 ps at longer mid-infrared wavelengths. The absence of a pronounced phonon bottleneck in the decay dynamics at room temperature suggests an efficient interplay of electron-electron and electron-phonon scattering even for photon energies below the optical phonon energy in graphene. The observed wavelength dependence of the photocurrent relaxation times is consistent with energy-dependent theoretical relaxation times as derived from a microscopic transport theory of graphene that includes electron-phonon coupling within a Holstein-Peierls Hamiltonian.

We investigate the interplay between frequency heterogeneity and higher-order triadic interactions in a ring network of Kuramoto oscillators. While both factors individually disrupt ordered states, their combination produces unexpected collective behavior. In the strong triadic coupling regime, moderate frequency heterogeneity substantially increases the global order parameter, with an optimal heterogeneity strength growing approximately linearly with triadic coupling strength. Basin stability analysis reveals that this order-promoting effect arises from a global restructuring of the attractor landscape: frequency heterogeneity shifts the attractor competition in favor of more ordered configurations. Linear stability analysis of frequency-locked twisted states reveals a competing effect: frequency heterogeneity monotonically erodes linear stability and reduces the probability of frequency locking. These two competing mechanisms, basin enlargement and linear destabilization, together account for the non-monotonic dependence of the order parameter on heterogeneity strength. Our results demonstrate that frequency heterogeneity can play a constructive role in oscillator networks with higher-order interactions.

The Boussinesq-Klein-Gordon (BKG) equation has emerged in the studies of nonlinear bulk strain waves in layered solid waveguides. The developed bi-directional weakly-nonlinear solution leads to two copies of the Ostrovsky equation, for the right- and left-propagating waves. Importantly, the derivation avoids the so-called `zero-mean contradiction' between the type of initial conditions in the parent equation and in the reduced model. In this paper, we apply the solution to describe the evolution of cnoidal waves with local periodicity defects and generic localised perturbations, and compare the results with the direct numerical simulations for the full BKG equation. The cnoidal waves with the periodicity defects discussed in our work constitute generalised travelling waves of the Korteweg-de Vries equation, while the Ostrovsky equation leads to a strong burst (and may lead to a rogue wave), qualitatively similar to the wavepacket emerging from a soliton initial condition, but appearing much faster. We compare the weakly-nonlinear solution with the direct numerical simulations within the bi-directional setting of the BKG equation and show that the discussed uni-directional waves and evolution scenarios remain stable in the presence of counter-propagating perturbations.

Tipping refers to the transition of a system from one state to another. In this study, we focus on shock-induced tipping, which occurs due to a sudden and large disturbance in a control parameter, which is referred to as the shock. This shock drives the system from one dynamical state to another. We present the first experimental demonstration of shock-induced tipping using a prototypical thermoacoustic system, the horizontal Rijke tube. In a thermoacoustic system, unsteady heat release and sound waves interact through positive feedback, leading to self-sustained, high-amplitude oscillations known as limit cycles. The system transitions from a quiescent state to a state of self-sustained oscillations when a shock is introduced in the power supplied to the heat source (an electrically heated grid). This shock is created by abruptly increasing the voltage supplied to the grid, which takes the system into a bistable region. To explain the underlying mechanism linking the shock in the supplied power to the observed tipping behaviour, we model the system by modifying the governing equations of the Rijke tube to incorporate the heat transfer properties of the grid. We demonstrate that the shock in the supplied power manifests as a shock in the grid temperature, causing the system to fall into the basin of attraction of an alternate stable state. The tipping event depends on the magnitude of the shock and the temperature of the grid. Understanding the mechanisms underlying shock-induced tipping is crucial for developing systems with improved safety and reliability.

Identifying the nodes that must be directly controlled to steer a network along a desired trajectory remains an open problem for digraphs, and even more so for hypergraphs. In this manuscript, we investigate network systems coupled via directed hypergraphs and consider a broad class of individual dynamics and coupling configurations, extending the definition of type II networks originally formulated for digraphs. For this class of networks with higher-order interactions, we establish necessary and sufficient conditions under which a pinning selection locally ensures successful control. Building on these analytical results, we propose a greedy heuristic for pinning control selection, which demonstrably outperforms existing methods.

Understanding how network structure gives rise to neuronal dynamics and whether compact computational models can recover that structure from data alone is a central challenge in computational neuroscience. We apply the performance-dependent network evolution (PDNE) framework to model the dynamics of the Wilson-Cowan (WC) neuronal system, a canonical two-population model of excitatory-inhibitory (E-I) interaction underlying physiological rhythms. Starting from a minimal seed network, PDNE iteratively grows and prunes a reservoir computing (RC) network based solely on prediction performance, yielding compact, task-optimized reservoirs networks. The evolved networks accurately predict both excitatory $E(t)$ and inhibitory $I(t)$ population activities across unseen stimulus amplitudes and generalize in a zero-shot manner to novel stimulus configurations: varying pulse number, position and amplitude without retraining. Structural analysis of the evolved networks reveals a consistent functional organization with nodes specialized for E, I, and shared E-I representations. Importantly, the population-level connectivity of the evolved reservoirs spontaneously recovers the correct excitatory-inhibitory sign pattern of the WC model for three of four interaction types, without this being imposed by design. These results demonstrate that performance-driven network evolution can produce not only accurate but structurally interpretable models of physiological rhythms, opening a path toward compact, data-efficient digital twins of neuronal systems.

Microring resonators (MRRs) are indispensable for wavelength filtering, sensing, and on-chip signal routing in photonic integrated circuits, yet visible-wavelength implementations using environmentally benign materials remain scarce. We report a numerical design study of add-drop MRRs employing Cs$_3$Cu$_2$Cl$_5$, a lead-free all-inorganic halide with favorable optical properties in the visible spectral range. Wavelength-resolved refractive index (n) and extinction coefficient (k) of Cs$_3$Cu$_2$Cl$_5$, calculated using density functional theory (DFT), are used as direct inputs to three-dimensional finite-difference time-domain (FDTD) simulations. Independent parametric sweeps are performed over ring waveguide width (500-900 nm), coupling gap (150-300 nm), and bend radius (5-20 um). At the balanced operating point of 600 nm ring width, 200 nm gap, and 10 um radius, the device achieves a loaded quality factor Q approx 5386, a free spectral range of 11.3 nm, a drop-port extinction ratio of 32.2 dB, and a finesse of 95.8. The coupling-gap sweep reveals the full transition from over-coupled through critically coupled to under-coupled operation, with the critical point occurring near 200 nm. A pronounced bending-loss threshold is observed between 5 and 10 um, below which all performance metrics degrade rapidly. These results provide the first systematic geometry-performance map for Cs$_3$Cu$_2$Cl$_5$ based microring resonators. Cross-platform validation using Tidy3D reproduces the spectral characteristics of the optimized device, and inverse design of the bus coupling region yields an additional 3 percent improvement in drop-port power transfer.

We present a fully reproducible demonstration of an AI-assisted scientific workflow designed for a broad physics, mathematics, and computer-science readership. The initial project artifact stack was generated from one single user prompt and then reviewed and curated for submission by the human author. Rather than claiming a new scientific discovery, the manuscript uses canonical benchmark problems with exact, manufactured, or independently checkable answers. The analytical component starts from the one-dimensional quantum harmonic oscillator, derives its dimensionless form, and validates finite-difference eigenpairs against exact Hermite-function benchmarks. The numerical partial-differential-equation component solves a heat equation with a known modal solution and a Poisson problem verified by a manufactured solution, with explicit convergence studies. The inverse-modeling component fits synthetic damped-oscillation data by nonlinear least squares and quantifies parametric uncertainty by bootstrap resampling. The computational-science component compares dense and sparse eigensolvers and contrasts direct and iterative sparse linear solvers, with careful interpretation of machine-dependent timing data. Taken together, the results show that contemporary AI can already serve as a useful scientific copilot for derivation, implementation, validation, visualization, and manuscript preparation, provided that each stage is constrained by benchmark theory, explicit verification, and transparent artifacts. The demonstration is therefore relevant not because the underlying science is novel, but because it offers a concrete template for trustworthy AI use in technical research practice.

First-principles calculations can accurately describe electron-phonon (e-ph) interactions and electronic transport in a wide range of materials, but are currently limited to unit cells with up to $\sim$100 atoms due to computational cost. Here, we develop an atomistic electronic potential with Holstein- and Peierls-like terms for modeling e-ph interactions and phonon-limited electronic transport that enables the study of moiré systems with thousands of atoms per unit cell. This method can accurately reproduce first-principles e-ph coupling and resistivity in graphene and large-angle twisted bilayer graphene (TBG). Using this approach, we study TBG over a range of twist angles down to 1.6$^\circ$ (5044-atom unit cell), and report the evolution of e-ph interactions and phonon-limited resistivity with twist angle. The predicted resistivity increases by two orders of magnitude between 13.2$^\circ$ and 1.6$^\circ$, driven by the progressive reduction of the electronic energy scale. Our calculations can predict key experimental trends in 2.0$^\circ$ and 1.6$^\circ$ TBG, including the resistivity and its dependence on temperature and band filling. Our work establishes a scalable approach for quantitative studies of e-ph interactions and transport in moiré materials and other systems with previously inaccessible length scales.

Spontaneous parametric downconversion (SPDC) has been considered as a reliable source of high- dimensional entangled states in orbital angular momentum (OAM) basis. In real-world experiments, the spatial walk-off of the pump often degrades the fidelity of the generated quantum state. Since the walk-off effect breaks the rotational symmetry of the system, the conservation of total OAM is violated. Although the compensation of walk-off effects has become a well-established experimental technique, a systematic modal analysis of the spatial walk-off effect is still incomplete for SPDC. Here, we quantitatively analyze the violation of OAM conservation due to the pump walk-off effect in SPDC processes. We have derived a scaling law of the total OAM distribution with respect to the pump walk-off angle. We have also explored the feasibility of using the spatial walk-off as a mechanism to engineer the generated quantum state. Our study has provided guidelines for the generation of OAM-entangled state under realistic experimental conditions.

Nitrogen-vacancy (NV) centers in diamond are a versatile quantum sensing platform for high sensitivity measurements of magnetic fields, temperature and strain with nanoscale spatial resolution. A common bottleneck is the analysis of optically detected magnetic resonance (ODMR) spectra, where target quantities are encoded in resonance features. Conventional nonlinear fitting is often computationally expensive, sensitive to initialization, and prone to failure at low signal-to-noise ratio (SNR). Here we introduce a robust, efficient machine learning (ML) framework for real-time ODMR analysis based on a one-dimensional convolutional neural network (1D-CNN). The model performs direct parameter inference without initial guesses or iterative optimization, and is naturally parallelizable on graphics processing units (GPU) for high-throughput processing. We validate the approach on both synthetic and experimental datasets, showing improved throughput, accuracy and robustness than standard nonlinear fitting, with the largest gains in the low-SNR regime. We further validate our methods in two representative sensing applications: diagnosing intracellular temperature changes using nanodiamond probes and widefield magnetic imaging of superconducting vortices in a high-temperature superconductor. This deep-learning inference framework enables fast and reliable extraction of physical parameters from complex ODMR data and provides a scalable route to real-time quantum sensing and imaging.

During the COVID-19 crisis, policymakers have implemented "social bubble" merging strategies, which allowed people from different households to meet and interact. Although these measures can mitigate the negative effects of extreme isolation, they also introduce additional contacts that may facilitate disease spread. As a result, several modeling studies have explored the epidemiological impact of different household-merging strategies, in which the selection of households to be merged is guided by specific demographic criteria, such as household size or the age composition of their members. Here we investigate an alternative pairing strategy in which households are merged according to the number of economically active (working) members. We develop a mathematical model of household networks using real demographic data from multiple regions around the world, and simulate a lockdown scenario in which only economically active individuals can leave their households, while the remaining non-working members stay indoors. By using numerical simulations and the generating function technique, we then estimate the epidemic risk for different household merging strategies. We found that merging strategies based on the number of working members can keep epidemic risk at similar levels as those based on household size. Moreover, the worker-based approach allows significantly more people to form larger social bubbles, exceeding 40\% of the population in some countries. We found that merging households with at most one worker provides the best balance between controlling epidemic risk and addressing people's need for social contact.

Despite the broadband nature of thermal light sources, optical spectroscopy over multiple spectral bands simultaneously remains challenging. Here, we demonstrate a practical Fourier transform infrared spectrometer (FTIR) that achieves room-temperature spectral coverage from 1 to 50 $μ$m (300--6 THz) in seconds using a single set of optics, with the long wave cutoff extendable to 90 $μ$m (3.3 THz) and the short wave to the ultraviolet (0.39 $μ$m). The interferometer employs a diamond plate beam splitter and windowless lithium tantalate (LTO) detector to probe the spectrum of combined incoherent sources operating at different temperatures. Applications of the instrument in modern chemometry, material science, and medicine are envisioned.

Detailed network models of social, biological and other complex systems are often dense, which increases their computational complexity in simulations and analysis. To address this challenge, graph sparsification is used to remove edges while preserving desired network properties. Distance backbones of weighted graphs, which remove edges that break a generalized triangle inequality for any given path-length measure, preserve all shortest paths of weighted graphs. They have been shown to typically sparsify graphs more, as well as preserve community structure and spreading dynamics better than alternative state-of-the-art methods. Here, We show that they significantly best preserve node centrality ranks, as well as local and global dynamics in spreading phenomena. This is done by introducing the distance backbone synthesis (DBS) to progressively sparsify weighted graphs according to a general family of nested distance backbones, whereby each edge is associated with the smallest distance backbone in which it appears. DBS provides a principled and natural method to sweep all degrees of sparsification possible while preserving connectivity, allowing us to precisely study (directed and undirected) weighted graph sparsification under multi-objective criteria. It provides an algebraically-principled explanation of edge importance by revealing the precise topological space associated with each edge. The theory is demonstrated with a battery of social contact networks obtained from real-world social activity in different scenarios. Our study also shows that the optimal preservation of node centrality and spreading dynamics happens for the distance backbone obeying the generalized triangle inequality for the path-length measure $g(x, y) = (\sqrt[3]{x}+\sqrt[3]{y})^3$, which removes more than half of the edges from the empirical networks studied.

Comparing multiple protein systems with variation such as different binding ligands or mutations, and understanding their effects is one of the objectives in molecular dynamics simulations. Representation of these systems by a few features enables quantitative comparison. However, because molecular dynamics simulation trajectories are high-dimensional spatiotemporal data, selection of key features relies on domain expertise, sometimes introducing arbitrary assumptions. Here, we present an approach that uses the optimal transport distance to compare high-dimensional trajectory data, and employs simulated annealing to identify the residues that best distinguish multiple systems. We term this algorithm auto-WHATMD (automated Wasserstein-based High-dimensional feature extraction Analysis for Trajectories of Molecular Dynamics). We applied auto-WHATMD to multiple protein-ligand systems of bromodomain 4 with different ligands, identifying the most discriminative residues in the loop region. Moreover, even a few selected residues were sufficient to capture the correlation with ligand-binding affinities, indicating that auto-WHATMD effectively prioritizes the most informative residues. Our approach can be used to efficiently determine key residues and design features for multiple analogous systems.

Advancing additive manufacturing, e.g., laser powder-bed fusion (LPBF), requires resolving rapid processes such as melt-pool dynamics and keyhole evolution in 4D (3D + time). Operando X-ray tomography is a state-of-the-art approach for 4D characterization, but its temporal resolution is fundamentally constrained by the sample rotation speed, limiting achievable 4D imaging rates and preventing the resolution of these fast phenomena. Here we present rotation-enabled X-ray Multi-Projection Imaging (rotation-XMPI), which captures three angularly resolved projections per time step and thereby decouples temporal resolution from the sample rotation speed. Combined with a self-supervised deep-learning reconstruction framework for multi-angle inputs, rotation-XMPI enables high-fidelity 4D imaging at unprecedented speed. We demonstrate the approach in an operando alumina laser-remelting experiment at MAX IV using three beamlets combined with 25 Hz sample rotation. Rotation-XMPI resolves melt-pool morphology and keyhole evolution; in contrast, conventional and limited-angle tomography remain rotation-limited, and motion blur prevents resolving these dynamics. Overall, rotation-XMPI delivers a 250-fold increase relative to state-of-the-art melt-pool imaging, effectively achieving 25,000 reconstructed volumes per second. This method establishes a practical route to scalable ultrafast 4D imaging for additive manufacturing and other materials processes.

Visible-light integrated photonics enables compact platforms for sensing, precision metrology, and free-space data links at visible wavelengths. However, many applications remain limited by the lack of high-speed and robust modulators in the blue-green band. Here we report, both operating at 532 nm, thin-film lithium tantalate waveguides of propagation losses of dB/cm scale and modulators with a flat frequency response to ~50 GHz. The modulator remains stable when delivering 5 dBm modulated optical power, which cannot be achieved by thin-film lithium niobate based counterparts under similar conditions and structures. We validate system-level underwater wireless optical communications (UWOCs) by transmitting 112 Gb/s signals over a 3-m underwater link. This represents the first integrated external modulator-based UWOC system, overcoming the bandwidth-power-chirp trade-offs of traditional directly modulated laser based systems. We further demonstrate dual-drive modulators for optical single-sideband and electro-optic frequency-comb generations in the green-wavelength band. These results provide a foundation for complex, robust, and active visible-light photonic integrated circuits for underwater optical applications.

Minimizing coating thermal noise is crucial for enhancing gravitational wave detector sensitivity, with a target Amplitude Spectral Density Reduction Factor (ASD RF) of $0.5$ relative to standard coatings. This study investigates the design of low-noise dielectric stacks using the 'Double Stack of Doublet' strategy, explored via ad-hoc optimization heuristics specifically developed for efficient parametric analysis of coating performance. We analyze the performance limits of ternary coatings based on SiO$_2$, Ti::SiO$_2$, and Ti::GeO$_2$, considering material property uncertainties and absorption constraints. Optimization results show that this system, even with relaxed absorbance constraint (1 ppm), falls short of the target, achieving a best ASD RF of $\sim 0.69$. Consequently, we explore alternative ternary 'Double Stack of Doublet' designs incorporating higher-refractive-index materials. Simulations demonstrate that incorporating alternative high-index materials offers a promising pathway, potentially enabling the achievement of the project target. We discuss the optimization strategies, performance trade-offs, design robustness, and implications of using high-index, potentially higher-loss materials for next-generation optical coatings.

This paper discusses the application of Dynamic Mode Decomposition (DMD) to the extraction of modal properties of linear mechanical systems, i.e., experimental modal analysis (EMA). First, theoretical background of the DMD is briefly reviewed and its relevance to the Ibrahim time-domain method is discussed. Second, DMD is applied to a single DOF system and multi-DOF discrete system to discuss the applicability and interpretation of the DMD as a method of EMA. Furthermore, the effects of measurement errors on the results of DMD are discussed. It is shown that with relatively small measurement errors, DMD can capture modal parameters accurately. However, with relatively large measurement errors, DMD fails to capture modal parameters. Finally, DMD is applied to experimentally-obtained displacement field of a cantilevered beam, and its modal parameters are extracted. It is shown that the modal parameters extracted by DMD are as accurate as the ones obtained by the existing modal parameter extraction method.

We establish a first-principles theory of vacuum Wannier functions unifying tight-binding and nearly-free-electron descriptions across solid-vacuum interfaces. Analytic solutions for canonical Wannier functions in arbitrary dimension and disentangled functions in 1D motivate a numerically verified 3D Wannier close-packing principle, enabling dense k-space construction of full Born-series scattering states at interfaces and thus predictive photoemission calculations without semiempirical vacuum potentials. Applications to graphene and h-BN reveal corrections beyond the first-Born approximation.

Rare nonadiabatic reactions are a key component of many important molecular processes but are challenging to capture with direct dynamical simulations. In this paper, we combine our recently developed mapping approach to surface hopping (MASH) with transition-path sampling to create a framework to efficiently simulate these rare events. This is possible because MASH trajectories are Markovian, time-reversible and obey Liouville's theorem. The combined approach generates nonadiabatic reactive pathways without biasing the underlying dynamics. The resulting ensemble allows for a detailed analysis of reaction mechanisms and the unraveling of statistical and dynamical properties, including rate constants. We apply the method to study a spin-boson model in thermal equilibrium over a wide range of diabatic coupling strengths. Our results demonstrate how this approach provides a practical and systematic tool for investigating rare nonadiabatic processes, potentially beyond the reach of brute-force simulations.

Optical chirality density is widely used as a scalar measure of the chiral properties of electromagnetic fields and their interaction with matter. However, in anisotropic and structured media, a single scalar quantity is generally insufficient to capture the full complexity of chiral field-matter coupling. In this work, we go beyond the conventional optical chirality density and introduce a set of tensor measures of electromagnetic chirality based on the Lipkin formalism. These tensor quantities provide a richer and more physically transparent description of chiral electromagnetic fields, particularly in an anisotropic environment. The physical meaning of individual tensor components is discussed, and their role in characterizing different aspects of electromagnetic chirality is clarified. The proposed approach reveals multiple, complementary measures of field chirality that naturally emerge in anisotropic cases and are directly relevant to the interaction of structured electromagnetic fields with matter.

The phenomenon of electron-laser vacuum breakdown is the multiple cascade production of electron-positron pairs in head-on collision of a beam of relativistic electrons with an intense laser pulse. This effect was first predicted by the author in 1996 [1] and further developed in [2]. In the present paper, an analytical expression for the total number of produced particles is obtained using the generalized Heitler model. The model results are shown to be in good agreement with the estimates of the pioneering works. An analysis of modern laser facilities (ELI, XCELS, European XFEL, Russian projects) is carried out and estimates of the expected effects are given. At ELI and XCELS class facilities, the quantum nonlinearity parameter can reach 60--150, corresponding to the deeply nonlinear QED regime with multiplicity up to 100 particles per seed electron. Experimental confirmation of the effect is expected in the coming years.

Optical microcombs generated in high-Q microresonators are promising chip-scale light sources for applications ranging from optical communications to spectroscopy and metrology. However, thermo-optic instabilities remain a major obstacle to reliable soliton access. Self-cooling using auxiliary modes can stabilize the intracavity power, yet part of the power is continuously allocated to thermal compensation rather than comb generation, thereby limiting comb power and bandwidth. Here we propose a thermal compensation scheme based on dynamic polarization control. During soliton initiation, a fraction of the pump is coupled to an orthogonally polarized mode to provide self-cooling and ensure reliable soliton access. After soliton formation, polarization rotation and pump tuning transfer this cooling power to the comb-generating mode, enabling efficient single-soliton operation. Using this approach, we experimentally demonstrate a broadband 108-GHz-FSR single-soliton microcomb spanning over 450 nm, together with approximately 39% improvement in the 20-dB bandwidth and 60% increase in comb power relative to the static self-cooling configuration. This dynamic polarization-based thermal compensation enables efficient use of available laser power and provides a practical route to high-performance soliton microcombs in platforms with strong thermo-optic effects.

We present the generalization of our FEM-based topology optimization framework to 3D blazed metasurfaces operating in reflection over the visible and near-infrared range [400-1,500]nm. The design region is described through a density-based SIMP interpolation and optimized using the adjoint method, enabling the treatment of several tens of thousands degrees of freedom. A first approach directly applies topology optimization to the 3D Finite Element mesh (mesh-based), yielding a freeform structure that achieves an average diffraction efficiency of 62% in order -1 over two octaves under the targeted incidence. However, such patterns remain difficult to manufacture. We therefore introduce a pillar-based parameterization, embedding fabrication constraints within the optimization loop. The resulting binary metasurface, compatible with e-beam lithography and Reactive Ion Etching techniques, achieves an average efficiency of 57% over the same spectral band in s-polarization, with low polarization dependence. This work demonstrates that large-scale 3D topology optimization can bridge the gap between broadband optical performance and realistic nanofabrication constraints for blazed metasurfaces.

Embedding nonabelian features into elastic metamaterials promises remarkable opportunities for wave control in many practical applications such as surface acoustic wave devices, mode multiplexers, and on-material computation. Nevertheless, current realizations are limited to arrangements of coupled resonators with fine-tuned interactions, limiting their applicability to continuous media. This theoretical and numerical study introduces a design principle for continuous nonabelian elastic metamaterial waveguides. The basic configuration consists of a composite waveguide made of multiple cylindrical waveguides coupled by spatially varying elements. These elements are engineered to follow geometrically-controlled parameter variations that cross selected degeneracies and produce a targeted nonabelian holonomy. The strategy based on crossing degeneracies fundamentally differs from abelian geometric phases, where parameters avoid and encircle degeneracies, or nonabelian Wilczek-Zee phases, where parameters are fine-tuned to maintain degeneracies throughout their cycle. The resulting holonomy transfers an input longitudinal excitation in one rod to an output response in another rod. When two such waveguides are concatenated, their ordering dictates the output response, thereby revealing the emergence of nonabelian dynamics. The nonabelian behavior persists across a broad range of frequencies and under perturbations to the geometry of coupling elements or cylinder diameters. These results establish a robust, effective, and practical route to leverage nonabelian physics in elastic metamaterials.

Grid-scale battery degradation unfolds over multi-year timescales under coupled electrochemical, thermal, and operational feedbacks difficult to capture using laboratory data or proprietary field datasets. This scarcity limits the development of degradation-aware algorithms and digital twins that require long-horizon, physically consistent ground truth. Here we present SAGE (Synthetic Aging for a Grid Environment), an open-source, physics-informed simulation framework that generates hour-resolved, multi-decade operating histories and degradation trajectories for heterogeneous battery energy storage system (BESS) fleets. The framework couples stochastic environmental drivers, market-based dispatch, electro-thermal behavior, aging kinetics, and asset-level heterogeneity within a transparent, externally parameterized architecture. We validate physical consistency through hierarchical tests, including Arrhenius temperature acceleration, thermal stratification, and emergent wear-out statistics. Simulations demonstrate how intrinsic heterogeneity in thermal environments and manufacturing naturally produces dispersion in state-of-health trajectories without imposed statistical failure assumptions. SAGE serves as a benchmarking platform for optimization, state estimation, and machine learning, enabling reproducible research in grid-scale energy storage modeling.

Two of the most common interpretations of quantum measurement disagree about the fate of quantum amplitudes after measurement, yet this disagreement has not previously led to experimentally distinguishable predictions. In the standard collapse picture, commonly linked to the Copenhagen interpretation of quantum mechanics, measurements eliminate unrealized amplitudes without leaving a memory. In contrast, in the unitary theory, the measurement device registers one of the possible outcomes while remaining part of an entangled state that continues to harbor the unrealized amplitudes. This persistence arises naturally under unitary evolution, since a measurement device that is part of an entangled system cannot serve as a faithful probe of the joint quantum state. Using single-photon measurements of a tunable quantum state, we experimentally show that these two theories make different predictions when three or more consecutive measurements are performed on the same quantum system. Analysis of the joint density matrix of the three measurements reveals coherence among them and supports the unitary theory of quantum measurement. When decoherence is explicitly introduced, the joint density matrix of the quantum system of interest and the apparatus becomes consistent with what a collapse theory would predict. This work clarifies the dynamics of consecutive quantum measurements and offers new insights into the interpretation of quantum measurements.

The blue shark (Prionace glauca) exhibits a striking dorsoventral color gradient, transitioning from vibrant blue dorsally to silver and white ventrally, a pattern widely interpreted as pelagic countershading. Despite its ecological significance, the physical basis of this coloration remains unresolved. Here we show that this color system does not arise from dermal chromatophores, as in most vertebrates, but from a previously unrecognised photonic architecture housed within the pulp cavity of individual dermal denticles that cover the skin. Optical imaging reveals discrete color domains within denticle crowns, while external denticle morphology remains similar across color zones. Using spectroscopy, micro-computed tomography, histology, and correlative electron microscopy, we demonstrate that color variation is organized across coupled micro- and nanoscale architectures. In blue denticles, iridophores and melanophores form a densely packed tessellated reflector-absorber system within an expanded crown-restricted pulp cavity. Transition-zone denticles exhibit partial cellular layering, whereas white denticles lack melanophores and contain only reflective cells. At the nanoscale, ordered purine-crystal stacks generate narrowband blue reflection, whereas disordered assemblies produce broadband white scattering. Together, these results reveal denticles as mechanically protected optical "pixels" whose hierarchical cellular and nanocrystal organization generates the shark's countershaded coloration.

Antiresonant hollow-core fibres (AR-HCFs) have recently reached attenuation far below the Rayleigh-scattering limit of silica, but their inherently multimode nature remains a major challenge for practical systems requiring high modal purity. In particular, suppressing higher-order modes (HOMs) at the 1 dB/m level while maintaining sub-0.1 dB/km fundamental-mode (FM) loss is difficult because conventional filtering strategies rely on tuning nested-tube dimensions, a design freedom that becomes increasingly restricted in the ultralow-loss regime. Here, we propose a new HOM-control mechanism in an interstitial-tube-assisted double nested anti-resonant nodeless fiber (IT-DNANF) by introducing angular offset of the interstitial tubes. Instead of using nested cavities as the primary tuning element, the proposed approach exploits the gap region between adjacent cladding tubes as a leakage-adjacent modal-control interface. Numerical simulations show that the offset increases both FM and HOM losses, but with a substantially stronger sensitivity for HOMs, leading to rapid enhancement of differential modal loss. Furthermore, when the gap-region FM is tuned into phase matching with the core HOM, strong coupling to a high-leakage state is induced, resulting in a pronounced HOM-loss peak. Using the practical criterion of HOM losslarger than 1 dB/m, we identify optimized IT-DNANF designs that achieve rapid HOM stripping while maintaining FM loss below 0.05 dB/km at 1550 nm. This work establishes angular offset as a physically distinct and manufacturability-friendly degree of freedom for mode purification in ultralow-loss hollow-core fibres.

Electromagnetic resonances play a central role in nanophotonics by enabling efficient confinement of electromagnetic energy and enhanced light-matter interaction. Traditionally, resonant phenomena have been described using platform-specific concepts developed within distinct research communities, including photonic crystals, plasmonics, and dielectric metasurfaces. In this Perspective, we propose a unified framework that distinguishes electromagnetic resonances as eigenmodes of open systems from their experimentally observed manifestations as scattering features. We show how resonances evolve from isolated particles to coupled oligomers and periodic structures, highlighting the roles of geometry, material response, and dimensionality. Particular attention is given to interference-driven phenomena such as bound states in the continuum, lattice resonances, anapoles, and superscattering, some of which cannot always be associated with a single eigenmode. By clarifying the relationship between eigenmodes, scattering channels, and interference effects, this Perspective provides a coherent language for interpreting resonant phenomena and identifies key challenges and opportunities for designing robust resonant photonic systems.

Rydberg-atom electric field sensing has shown great potential from near-DC to THz with state-of-the-art measurement metrics realized in sensitivity, phase extraction, multi-band receptivity, etc. While Rydberg-atom sensors have shown exceptional performance in the GHz regime, low-frequency operation has remained challenging because of electric-field-screening in conventional vapor cells, which suppresses externally applied fields. We overcome this limitation by combining auxiliary modulation and lock-in detection with a paraffin-coated vapor cell, and demonstrate an electrode-free, wideband method for sensing frequencies, ranging from 0.5 Hz to 10 kHz. Our work extends Rydberg-atom sensor range to VLF, ULF, SLF, ELF and sub-ELF frequency bands. In our method, high state-of-the-art sensitivities have been achieved - 819 $μ$V/cm/$\sqrt{\text{Hz}}$ for 1 Hz, 33 $μ$V/cm/$\sqrt{\text{Hz}}$ for 10 Hz, 10 $μ$V/cm/$\sqrt{\text{Hz}}$ for 100 Hz and 2 $μ$V/cm/$\sqrt{\text{Hz}}$ for 1 kHz.

Scaling solid-state architectures to the millions of qubits required for utility-scale quantum computing could benefit from the integration of control electronics in the immediate vicinity of the quantum layer. However, lithographically fabricated solid-state qubits perform best at temperatures well below 1 K, where available cooling power is limited, whereas the control electronics dissipate substantial power and therefore require the higher cooling power available at elevated temperatures. To address this challenge, we propose a cryopackaging concept that uses broadband phononic Distributed Bragg Reflectors (DBRs) as a thermal barrier between cryoelectronics and the qubit chip. As an experimental realization of this concept, we fabricate and characterize Ta/SiO$_2$ DBR structures. In this architecture, the DBR is intended to provide mechanical support for superconducting vias while offering substantially better thermal insulation than typical bulk materials. For a 600-nm-thick DBR consisting of 10 Ta/SiO$_2$ bilayers, we obtain a thermal conduction below 1 mW/cm$^2$ from 1.5 K to 100 mK. In a centimeter-scale architecture, this level of isolation is compatible with Watt-level cooling power for nearby electronics while maintaining a qubit temperature around 100 mK in commercially available dilution refrigerators.

A high-resolution Eulerian method for simulating high-speed polydisperse granular multiphase flows has been developed. The governing equations include a compressible gas that is coupled to mass-based moment equations for a polydisperse granular flow derived from the generalized population balance equation. The model includes effects from particle collisions, drag, convective heat transfer, particle-fluid-particle pressure, and finite-size particle force terms. The mass moment integrals are closed using the generalized quadrature method of moments to allow for continuous size distributions. The governing equations are solved by using high-resolution reconstruction schemes and results from decoupled Riemann problems for the gas and particles as each quadrature node. Success of the technique is demonstrated through a variety of numerical experiments including polydisperse multiphase Riemann shock-tube problems, shock--particle-curtain interactions, dust layer dispersal, dust layer dispersal by shock waves, and dispersal of spherical particle shells by high-pressure gas.

Quantum systems inside high-Q cavities offer an excellent testbed for the control of emergent symmetries induced by light and their interplay with quantum matter. Recently several developments in cavity experiments with neutral atoms and other quantum objects such as ions motivate the study of their quantum correlated properties and their entanglement to tailor and control the behavior of the system. Using the enhanced coupling between light and interacting matter we explore the properties of emergent superradiant modes using our newly developed Light-Matter DMRG algorithm with strongly interacting spin chains. We explore a experimentally viable generalization of the transverse Ising chain coupled to the cavity light where it is possible to induce multimode structures tailored by the light pumped into the system. We find a plethora of scenarios can be explored with clear and accesible measurable signatures. This allows to study the physics of emergent orders and strong quantum correlations with quantum spins where the local and long range coupling can be efficiently simulated. We find that quantum spin nematic states with long range order and magnon pairs emerge as the transitions to superradiant phases take place. Notably, we show the cavity field allows the optimization of entanglement between spins for different light induced modes which can be used for quantum state engineering of quantum correlated states. Our methods can be used to model other hybrid quantum systems efficiently.

Murray's cubic branching law ($α=3$) predicts a universal diameter scaling exponent for all hierarchical transport networks, yet arterial trees yield $α\sim 2.7-2.9$. We show that this discrepancy has a structural origin: Murray's universality is an artifact of cost homogeneity, not a biological property. Incorporating the empirical vessel-wall thickness law $h(r)=c_0 r^p$ ($p \approx 0.77$) introduces a third metabolic cost term $\propto r^{1+p}$ that renders the cost function inhomogeneous with incommensurate scaling exponents. By Cauchy's functional equation, homogeneity is necessary and sufficient for a universal branching exponent to exist; its absence implies non-universality, and Murray's law is identified as a singular degeneracy of the cost-function family rather than a general principle. We prove that the resulting scale-dependent exponent satisfies the strict bounds $(5+p)/2 < α^*(Q) < 3$ independently of flow asymmetry (Theorem 4, Corollary 5). The static wall-tissue mechanism bounds the symmetric bifurcation exponent to $α_t \in [2.90, 2.94]$ from measured parameters, marking a first-order symmetry breaking from Murray's law that narrows the empirical gap by one-third. The remaining discrepancy with the cardiovascular mean ($α_{exp} \approx 2.70$) is not a model failure but a mathematical necessity that signals the independent contribution of pulsatile wave dynamics. Additionally, the wall cost breaks Murray's topological degeneracy, bounding the optimal branching number to small finite integers; binary bifurcation emerges as the physiologically selected minimum under steric constraints.

Reduced graphene oxide (rGO) exhibits strong anisotropic light absorption and high compatibility with photonic integrated chips, making it a promising material for implementing high performance onchip polarization selective devices. The performance of rGO integrated waveguide polarizers is highly dependent on the waveguide geometry, and achieving optimal performance requires exploring a large parameter space, making conventional mode simulation methods computationally demanding. Here, we propose and demonstrate a machine learning framework based on fully connected neural networks (FCNNs) to map the dependence of the polarizer figure of merit (FOM) on the waveguide geometry. Once trained by using a small dataset of low resolution mode simulation results, the FCNN framework can rapidly and accurately predict FOM values across a large structural parameter space with high resolution. Results show that this method can reduce overall computing time by more than 4 orders of magnitude as compared to the mode simulation methods, and achieve high prediction accuracy with an average deviation (AD) below 0.05. These results highlight the FCNN based machine learning framework as an efficient tool for the design and optimization of rGO integrated waveguide polarizers.

Spin-orbit beams, in which the orbital angular momentum degree of freedom is coupled to a two-level system such as polarization of light or spin in electrons and neutrons, have gained significant interest for their unique propagation properties and potential applications in imaging, material characterization, optical trapping, and quantum information processing. In this work we introduce a method for generating and engineering two-dimensional lattices of such spin-orbit beams based on coherent averaging in k-space. By programming the angle, amplitude, and polarization of a set of input beams we obtain precise control over lattice geometry and period, as well as the orbital and radial degrees of freedom inside each unit cell. We explore both electromagnetic and matter wave implementations, and we experimentally demonstrate the generation and characterization of a micron-scale optical hexagonal lattice with well defined orbital and radial numbers in each unit cell. The described methods provide a robust and general method of generating and controlling structured waves such as optical skyrmions and matter wave implementations of orbit and spin-orbit beams.

Continuous-variable quantum key distribution (CV-QKD) enables secure communication over standard telecom infrastructure, yet its scaling is stalled by bulky, discrete optical hardware. We address this bottleneck by demonstrating a real-time CV-QKD system driven by a chip-scale hybrid transmitter built from commercial telecom components. By integrating a micro-optic external-cavity laser with a monolithic photonic integrated IQ modulator, we provide high performance, enabling secret-key generation over 102 km of optical fiber, while reducing the size of the optics by 95%. Moreover, real-time operation overcomes the offline post-processing bottlenecks of experimental setups. This work bridges laboratory demonstrations and field-deployable technology, with a scalable architecture for cost-effective quantum networks.

Carbapenem-resistant Pseudomonas aeruginosa (CRPA) and Acinetobacter baumannii (CRAB) represent a major clinical and epidemiological challenge and pose a growing threat to public health and the environment. Accordingly, CRPA and CRAB were investigated in hospital wastewater (HWW) collected during winter and summer 2024 from 64 healthcare facilities across all 16 Polish voivodeships. To our knowledge, this study constitutes the first nationwide, large-scale assessment in Poland of carbapenem resistance in these high-risk pathogens in hospital wastewater. The study aimed to determine the prevalence of carbapenem-resistant bacteria (CRB) in HWW discharged into the public sewer system and municipal wastewater treatment plants (WWTPs). In addition, associations between CRB prevalence, hospital geographic location, and sampling season were analyzed to identify spatial and temporal patterns of carbapenem resistance (CR). Carbapenem-resistant P. aeruginosa were predominant in all studied regions. Carbapenem-resistant A. baumannii were identified in a smaller percentage of samples and were characterized by greater genotypic diversity. The ERIC-PCR assay confirmed the presence of both closely related strains and unique genetic profiles, which suggests that CRB emissions into the environment have a complex character. The statistical analysis revealed significant relationships between CRB counts, the physicochemical parameters of HWW, and antibiotic concentrations in HWW samples. In addition, the tested samples harbored many antibiotic resistance genes (ARGs), which confirms that HWW is a significant reservoir of mobile genetic elements (MGEs) involved in the spread of antibiotic resistance. The results of the study indicate that HWW should be rigorously monitored and managed to minimize risks to public health and environment.

Gene regulatory networks exhibit hierarchical organization across scales; capturing this structure mathematically requires a metric that distinguishes regulatory influence at each level. We show that the ultrametric of the $p$-adic integers $\mathbb{Z}_p$ -- whose self-similar nested-ball structure is a natural fractal encoding of multi-scale organization -- provides such a framework. Embedding the $N$-gene state space into $\mathbb{Z}_p$ and working over the complete, algebraically closed field $\mathbb{C}_p$, we prove the existence of rational functions that interpret the discrete dynamics and construct hierarchical approximations at each resolution level. These constructions yield a stability measure $μ$ -- aggregating how the dynamics contracts or expands across resolution levels -- and a ball-level classification of fixed points -- contracting, expanding, or isometric -- extending the attracting/repelling/indifferent trichotomy of non-Archimedean dynamics from points to balls. A key result is that $μ$ and the classification, although their definition and dynamical meaning require the analytical tools of $\mathbb{C}_p$, are fully determined by the discrete data. Minimizing $μ$ over all $N!$ gene orderings defines an optimal regulatory hierarchy; for the Arabidopsis thaliana floral development network ($N=13$, $p=2$), a $μ$-minimizing ordering places known master regulators -- UFO, EMF1, LFY, TFL1 -- in the leading positions and recovers the accepted developmental hierarchy without biological input beyond the transition map.

Objectively verifying the generative mechanism of consciousness is extremely difficult because of its subjective nature. As long as theories of consciousness focus solely on its generative mechanism, developing a theory remains challenging. We believe that broadening the theoretical scope and enhancing theoretical unification are necessary to establish a theory of consciousness. This study proposes seven questions that theories of consciousness should address: phenomena, self, causation, state, function, contents, and universality. The questions were designed to examine the functional aspects of consciousness and its applicability to system design. Next, we will examine how our proposed Dual-Laws Model (DLM) can address these questions. Based on our theory, we anticipate two unique features of a conscious system: autonomy in constructing its own goals and cognitive decoupling from external stimuli. We contend that systems with these capabilities differ fundamentally from machines that merely follow human instructions. This makes a design theory that enables high moral behavior indispensable.

This note presents a concise mathematical formulation of tightly-coupled LiDAR-Inertial Odometry within an iterated error-state Kalman filter framework using a VoxelMap representation. Rather than proposing a new algorithm, it provides a clear and self-contained derivation that unifies the geometric modeling and probabilistic state estimation through consistent notation and explicit formulations. The document is intended to serve both as a technical reference and as an accessible entry point for a foundational understanding of the system architecture and estimation principles.

Accurate channel state information (CSI) prediction is crucial for next-generation multiple-input multiple-output (MIMO) communication systems. Classical prediction methods often become inefficient for high-dimensional and rapidly time-varying channels. To improve prediction efficiency, it is essential to exploit the inherent low-rank tensor structure of the MIMO channel. Motivated by this observation, we propose a dynamic mode decomposition (DMD)-based prediction framework operating on the low-dimensional core tensors obtained via a Tucker decomposition. The proposed method predicts reduced-order channel cores, significantly lowering computational complexity. Simulation results demonstrate that the proposed approach preserves the dominant channel dynamics and achieves high prediction accuracy.

Affine frequency division multiplexing (AFDM) and orthogonal time frequency space (OTFS) are two promising advanced waveforms proposed for reliable communications in high-mobility scenarios. In this paper, we introduce a simple transmit diversity technique, termed cyclic delay-Doppler shift (CDDS), for these two advanced waveforms to achieve ultra-reliable communications in doubly selective channels (DSCs). Two simple CDDS schemes, named modulation-domain CDDS (MD-CDDS) and time-domain CDDS (TD-CDDS), are proposed, which perform CDDS in advance at the transmitter before and after the modulation, respectively. We demonstrate that both of the two proposed CDDS schemes can be implemented efficiently and flexibly by multiplying the transmit vector with a well-designed precoding matrix, which is nothing but a sparse phase-compensated permutation matrix. Moreover, we theoretically and numerically prove that CDDS can provide MIMO-AFDM and MIMO-OTFS with optimal transmit diversity gain when a proper CDDS step is adopted. Compared to the conventional transmit diversity techniques, the proposed CDDS scheme enjoys the advantages of lower channel estimation overhead, implementation complexity, and signal processing latency, making it particularly suitable for ultra-reliable communications in high-mobility scenarios.

Programmable wireless environments (PWEs), empowered by reconfigurable intelligent surfaces (RISes), have emerged as a transformative paradigm for next-generation networks, enabling deterministic control over electromagnetic (EM) propagation to enhance both performance and security. In this work, we introduce RF-Fencing, a novel RIS-enabled PWE service that enforces spatially selective control over wireless transmissions, simultaneously suppressing unwanted signal exposure while sustaining robust connectivity for legitimate users. To realize this vision, we develop SHIELD, a lightweight and scalable algorithm that orchestrates multiple RIS units by multiplexing precompiled codebook entries with real-time, low-complexity optimization. Through extensive evaluations across diverse frequencies, RIS configurations, and deployment scenarios, SHIELD demonstrates both far-field directional control and near-field quiet-zone creation, thereby enhancing network security. Our findings reveal that SHIELD effectively balances proactive covert communication with service delivery by dynamically managing multiple signal suppression and delivery areas, while enabling the realization of EM quiet zones with minimal impact on surrounding regions, ultimately establishing RF-Fencing as a practical RIS-based foundation for privacy-preserving and adaptive wireless environments in future 6G networks.

Channel models are essential for the design, evaluation, and optimization of wireless communication systems. The emerging space-air-ground-sea integrated network (SAGSIN), characterized by diverse service applications and extended-spectrum operations, places even greater demands on highly accurate channel models. However, conventional channel sounding is limited by generalized measurement campaigns, inadequate cross-band consistency, and insufficient real-time adaptability, making it unable to meet the needs of SAGSIN for scenario-specific and high-precision channel modeling. To address this challenge, we propose a novel technological framework, termed integrated channel sounding and communication (ICSC). By deeply integrating sounding and communication, the ICSC enables efficient and real-time acquisition of dynamic channel characteristics during communication processes, supporting fine-grained site- and scenario-specific measurements. Furthermore, leveraging artificial intelligence techniques, ICSC can identify channel conditions and adapt waveform parameters in real-time according to scenario variations, which in turn enhances communication performance. This article first introduces the fundamental principles of the ICSC framework, elaborates on its core concepts and key advantages, and demonstrates its feasibility through the development of an integrated verification system (IVS). Subsequently, the potential applications and opportunities of the ICSC are analyzed in depth, followed by a discussion of its future development directions and remaining challenges.

Accurate delay-Doppler channel estimation is critical for next-evolution waveforms (NEWs) to enable reliable signal detection. This paper proposes a robust channel estimation algorithm that integrates Flag sequences optimized via an adaptive accelerated parallel majorization-minimization (AP-MM) algorithm with a proposed channel estimation algorithm. To enable efficient, low-complexity parameter extraction and further overcome the robustness issues of conventional greedy estimation, we introduce two key enhancements, i.e., a candidate selection strategy to mitigate spurious sidelobe peaks, and a global least squares (LS) refinement stage to eliminate error propagation caused by sidelobe masking effects. Numerical results demonstrate that the proposed scheme significantly outperforms traditional existing algorithms, achieving the desired estimation accuracy.

A target recognition framework relying on near-field integrated sensing and communication (ISAC) systems is proposed. By exploiting the distance-dependent spatial signatures provided by the near-field spherical wavefront, high-accuracy sensing is realized in a bandwidth-efficient manner. A spatio--temporal--frequency (STF) transformer framework is introduced for target recognition using electromagnetic features found in the wireless channel response. In particular, a lightweight spatial encoder is employed to extract features from the antenna array for each frame and subcarrier. These features are then fused by a time-frequency transformer head with positional embeddings to model temporal dynamics and cross-subcarrier correlations. Simulation results demonstrate that strong target recognition performance can be achieved even with limited bandwidth resources.

This paper investigates physical-layer security (PLS) enabled by graph neural networks (GNNs). We propose a two-stage heterogeneous GNN (HGNN) to maximize the secrecy energy efficiency (SEE) of a reconfigurable intelligent surface (RIS)-assisted multi-input-single-output (MISO) system that serves multiple legitimate users (LUs) and eavesdroppers (Eves). The first stage formulates the system as a bipartite graph involving three types of nodes-RIS reflecting elements, LUs, and Eves-with the goal of generating the RIS phase shift matrix. The second stage models the system as a fully connected graph with two types of nodes (LUs and Eves), aiming to produce beamforming and artificial noise (AN) vectors. Both stages adopt an HGNN integrated with a multi-head attention mechanism, and the second stage incorporates two output methods: beam-direct and model-based approaches. The two-stage HGNN is trained in an unsupervised manner and designed to scale with the number of RIS reflecting elements, LUs, and Eves. Numerical results demonstrate that the proposed two-stage HGNN outperforms state-of-the-art GNNs in RIS-aided PLS scenarios. Compared with convex optimization algorithms, it reduces the average running time by three orders of magnitude with a performance loss of less than $4\%$. Additionally, the scalability of the two-stage HGNN is validated through extensive simulations.

Modern radars employing wideband signals and extremely large (XL) multiple-input multiple-output (MIMO) arrays can significantly improve range and angular resolution. However, when large bandwidth and array aperture are used simultaneously, the spatial delay across the array becomes comparable to the radar range resolution, leading to the spatial wideband effect (SWE). The SWE introduces several distortions including range migration (range squint), beam squint, and range-angle coupling (RAC), which spread the target response in the range-angle domain and may cause physically separated targets to overlap and mask each other. In this work, we propose a decoupling-based target detection and parameter estimation framework for MIMO frequency modulated continuous wave (FMCW) radar. The proposed method reformulates the joint range-angle estimation problem as a decoupled sequential frequency estimation problem, where the two-dimensional (2D) estimation is carried out through successive one-dimensional (1D) super-resolution estimations. Specifically, we employ orthogonal matching pursuit (OMP) to perform sparse recovery-based range and angle estimation with high resolution. The proposed decoupling strategy is further extended to spatial wideband XL-MIMO FMCW radar systems, enabling reliable detection and separation of targets even when their responses overlap due to severe RAC. Simulation results demonstrate that the proposed approach accurately detects multiple targets and successfully resolves overlapping target responses in the presence of SWE, outperforming conventional Fourier transform and clustering-based methods.

Extremely large antenna arrays (ELAAs) are widely adopted in mmWave/THz communications to compensate for the severe path loss, wherein the channel estimation remains a significant challenge since the Rayleigh distance of ELAAs stretches to tens or even hundreds of meters and the near-field channel model should be considered. Existing polar-domain based methods and block-sparse based methods are originally devised for Uniform Linear Arrays (ULAs) near-field channel estimation. The polar-domain based method can be applied to Uniform Planar Arrays (UPAs), but it behaves plain since it ignores the specific sparsity structure of the UPA near-field channels. Meanwhile, the block-sparse based method cannot be extended to the UPA scenarios directly. To address these issues, we first reformulate the original UPA near-field channel as an outer product of two ULA near-field channels and we construct a modified two-dimensional DFT (2D-DFT) dictionary for it. With the proposed dictionary, we further prove that the UPA near-field channel admits a 2D block-sparse structure. Leveraging this specific sparse structure, we solve the channel estimation problem with the 2D Pattern-Coupled Sparse Bayesian Learning (2D-PCSBL) algorithm. Simulation results show that the proposed approach outperforms conventional existing methods while maintaining a comparable computational complexity.

Delay-coordinates dynamic mode decomposition (DC-DMD) is widely used to extract coherent spatiotemporal modes from high-dimensional time series. A central challenge is distinguishing dynamically meaningful modes from spurious modes induced by noise and order overestimation. We show that model order detection and mode selection in DC-DMD are fundamentally problems of subspace geometry. Specifically, true modes are characterized by concentration within a low-dimensional signal subspace, whereas spurious modes necessarily retain non-negligible components outside any moderate overestimate of that subspace. This geometric distinction yields a perturbation-robust definition of true and spurious modes and yields fully data-driven selection criteria. This geometric framework leads to two complementary data-driven selection criteria. The first is derived directly from the geometric distinction and uses a data-driven proxy of the signal-subspace to compute a residual score. The second arises from a new operator-theoretic analysis of delay embedding. Using a block-companion formulation, we show that all modes exhibit a Kronecker-Vandermonde (KV) structure induced by the delay-coordinates, and true modes are distinguished by the degree to which they conform to it. Importantly, we also show that this deviation is governed precisely by the geometric residual. In addition, our analysis provides a principled explanation for the empirical behavior of magnitude- and norm-based heuristics, clarifying when and why they fail under delay-coordinates. Extensive numerical experiments confirm the theoretical predictions and demonstrate that the proposed geometric and structure-based methods achieve robust and accurate order detection and mode selection, consistently better than existing baselines across noise levels, spectral separations, damping regimes, and embedding lengths.

In this paper, a unified framework for chirp-domain waveforms, including orthogonal chirp division multiplexing (OCDM) and affine frequency division multiplexing (AFDM), is developed. Based on their continuous-time representations, we show that these waveforms fall within the conventional Weyl-Heisenberg (WH) framework for multicarrier (MC) waveforms, where the root chirp corresponds directly to the prototype pulse in the WH framework. Since the chirp is a constant-envelope signal and is transparent to subcarrier orthogonality, these waveforms can be further interpreted as pulse-shaped (PS) orthogonal frequency division multiplexing (OFDM). Within the developed PS-OFDM framework, the power spectral density of chirp-domain waveforms is derived analytically. We then discuss existing practical implementations of chirp-domain waveforms, which rely on sub-Nyquist discrete-time samples and therefore exhibit frequency aliasing. The resulting aliased waveform is analyzed, and the orthogonality among the embedded aliased chirps is discussed. It is shown that the aliased chirps are conditionally orthogonal, whereas the implemented approximate aliased chirps can maintain mutual orthogonality when an appropriate sample-wise pulse-shaping filter is applied. We further derive an exact input-output relation for the implemented chirp-domain waveform over a delay-Doppler (DD) channel, showing that the effective channel observed at a practical receiver does not, in general, admit a DD spreading-function model commonly assumed in the literature. The implementation complexity is also investigated and compared with that of orthogonal delay-Doppler division multiplexing (ODDM), the DD-domain MC waveform defined within the evolved WH framework. Finally, simulation results are provided to verify the analysis.

Integrated sensing and communications (ISAC) has been envisioned as a promising solution to support emerging services in low-altitude wireless networks (LAWNs), where upgrading 5G ground base stations (GBS) toward new active sensing systems with wide coverage, low cost, high accuracy, and favorable spectrum compatibility, is strongly desired. However, such an evolution faces several critical challenges, particularly in the detection and tracking of weak and slow unmanned aerial vehicles (UAVs). These challenges include ISAC waveform design, clutter cancellation resilient to high clutter-to-noise ratios (CNRs), and efficient Doppler separation between UAVs and clutter. To that end, we summarize potential solutions and raise a comprehensive framework on implementing the 5Gadvanced (5G-A) GBS. Outfield experiments demonstrate that the developed 5G-A GBS can effectively track weak and slow targets at distances exceeding 1 kilometer, while incurring only a 1.2% downlink rate loss relative to commercial 5G-A GBS.

The transformative power of artificial intelligence (AI) and machine learning (ML) is recognized as a key enabler for sixth generation (6G) mobile networks by both academia and industry. Research on AI/ML in mobile networks has been ongoing for years, and the 3rd generation partnership project (3GPP) launched standardization efforts to integrate AI into mobile networks. However, a comprehensive review of the current status and challenges of the standardization of AI/ML for mobile networks is still missing. To this end, we provided a comprehensive review of the standardization efforts by 3GPP on AI/ML for mobile networks. This includes an overview of the general AI/ML framework, representative use cases (i.e., CSI feedback, beam management and positioning), and corresponding evaluation matrices. We emphasized the key research challenges on dataset preparation, generalization evaluation and baseline AI/ML models selection. Using CSI feedback as a case study, given the test dataset 2, we demonstrated that the pre-training-fine-tuning paradigm (i.e., pre-training using dataset 1 and fine-tuning using dataset 2) outperforms training on dataset 2. Moreover, we observed the highest performance enhancements in Transformer-based models through fine-tuning, showing its great generalization potential at large floating-point operations (FLOPs). Finally, we outlined future research directions for the application of AI/ML in mobile networks.

Stochastic Optimal Control Problems (SOCPs) plays a major role in the sequential decision-making challenges. There exist various iterative algorithms, under framework of stochastic maximum principle, that sequentially find the optimal control decision. However, they are based on the adjoint sensitivity analysis that necessitates simulation of an adjoint process, typically a backward stochastic differential equation (SDE) that must simultaneously be adapted to a forward filtration and satisfy a terminal condition, which substantially increases complexity and exacerbates the curse of dimensionality. We instead develop a stochastic maximum principle based on the Malliavin calculus, which enables us to devise an iterative algorithm without need of an adjoint process. Our algorithm however needs the Malliavin derivative that can be efficiently computed based on a forward simulator. Empirical comparisons against standard iterative algorithms demonstrate that our approach alleviates the dimensionality bottleneck while delivering competitive performance on the considered SOCPs.

Stacked intelligent metasurfaces (SIMs) extend the concept of reconfigurable intelligent surfaces by cascading multiple programmable layers, enabling advanced electromagnetic wave transformations for communication and sensing applications. However, most existing optimization frameworks rely on simplified channel abstractions that may overlook key electromagnetic effects such as multiport coupling, circuit losses, and non-ideal hardware behavior. In this paper, we develop a modeling and optimization framework for SIMs based on a multiport network representation using scattering parameters. The proposed formulation captures realistic circuit characteristics and mutual interactions among SIM ports while remaining amenable to optimization. The resulting models are validated through electromagnetic simulations, enabling a systematic comparison between idealized and practical SIM configurations. Numerical results for communication and sensing scenarios confirm that the proposed framework provides accurate performance predictions and enables the effective design of SIM configurations under realistic electromagnetic conditions.

Large-scale MIMO detection remains challenging because exact or near-maximum-likelihood search is difficult to scale, while available quantum resources are insufficient for directly solving full-size detection instances by QAOA. This paper therefore proposes a Block-QAOA-Aware MIMO Detector (BQA-MD), whose primary purpose is to reorganize the detection chain so that it becomes compatible with limited-qubit local quantum subproblems. Specifically, BQA-MD combines block-QAOA-aware preprocessing in the QR domain, a standards-consistent blockwise 5G NR Gray-HUBO interface, an MMSE-induced dynamic regularized blockwise objective, and K-best candidate propagation. Within this framework, fixed-size block construction gives every local subproblem a uniform circuit width and parameter dimension, which in turn enables parameter-transfer QAOA as a practical realization strategy for structurally matched local subproblems. Experiments are conducted on a 16x16 Rayleigh MIMO system with 16QAM using classical simulation of the quantum subroutine. The results show that the regularized blockwise detector improves upon its unregularized counterpart, validating the adopted blockwise objective and the block-QAOA-aware design rationale. They also show that the parameter-transfer QAOA detector nearly matches the regularized blockwise exhaustive reference and clearly outperforms direct-training QAOA in BER, thereby supporting parameter reuse as the preferred QAOA realization strategy within the proposed framework. In the tested setting, MMSE remains slightly better in the low-SNR region, whereas the parameter-transfer QAOA detector becomes highly competitive from the medium-SNR regime onward.

Integrated Sensing and Communications (ISAC) has emerged as a key enabler for sixth generation (6G) wireless systems by jointly supporting data transmission and environmental awareness within a unified framework. However, communication and sensing functionalities impose inherently conflicting performance requirements, particularly in heterogeneous networks where users may demand sensing only, communication only, or joint services. Selecting a waveform that satisfies diverse service demands therefore becomes a challenging multi objective decision problem. In this paper, a multi objective learning approach for adaptive waveform selection in ISAC systems is proposed. A simulation driven evaluation framework is developed to assess multiple waveform candidates across communication, sensing, and joint performance metrics. Instead of enforcing scalar utility aggregation, waveform performance is represented in a multi dimensional objective space where Pareto optimal candidates are identified for each scenario. A dataset is generated by varying user demand distributions and channel conditions, and multi-label targets are constructed based on Pareto dominance. Machine learning models are trained to learn the mapping between network conditions and Pareto optimal waveform sets, enabling fast waveform selection under dynamic network states. Simulation results demonstrate that the proposed framework effectively adapts waveform selection to heterogeneous service requirements while preserving sensing communication trade offs, providing a forward-looking perspective for 6G and beyond ISAC deployments.

Distributed multiple-input multiple-output (MIMO) architectures enable large-scale integrated sensing and communication (ISAC) by providing high spatial resolution and robustness through spatial diversity. However, practical phase-coherent sensing is challenged by phase synchronization errors and modeling mismatch caused by grid discretization. Existing over-the-air (OTA) synchronization methods typically treat synchronization and sensing tasks separately, which may lead to inaccurate phase alignment when multipath components are used for imaging. In this paper, we propose a non-line-of-sight (NLOS)-aided joint OTA synchronization and off-grid imaging framework for distributed MIMO ISAC systems. First, a line-of-sight (LOS)-assisted coarse synchronization is performed to establish initial phase coherence across distributed links. Subsequently, an iterative refinement stage exploits reconstructed NLOS components obtained from imaging results. By modeling off-grid effects via a first-order Taylor expansion, we transform measurements with nonlinear off-grid offset into an augmented linear model with jointly sparse reflectivity and off-set variables. The imaging problem is reformulated as a structured sparse recovery task and solved using a tailored off-grid approximate message passing (OG-AMP) algorithm. The imaging and synchronization modules are coupled within a closed-loop alternative optimization framework, where improved imaging enables more accurate phase refinement, and vice versa. Numerical results show that the proposed framework achieves accurate synchronization and imaging under phase errors. Compared with conventional approaches, it shows superior robustness and accuracy.

Pinching-antenna systems (PASs) have been proposed as a flexible antenna technology to fulfill the stringent requirements of high data rate and large-scale equipment deployment in future wireless networks. The principle of PA involves mapping a signal over dielectric waveguides for transmission. By adjusting the positions of pinching antennas (PAs) over the waveguides, with the aim of gain enhancement for line-of-sight links and the reduction of large-scale path loss. Symbol-level precoding (SLP) is a nonlinear precoding technique, which converts multi-user interference into constructive interference via beamforming design at symbol level. In this paper, we study the combination of SLP and PAS, leveraging the advantages of PAS to further enhance the ability of SLP to convert constructive interference. The transmit power minimization problem is formulated and solved for the multiple waveguides multiple PAs system by jointly beamforming and PAs' positions design under the SLP principle. The alternating optimization (AO) framework is applied to decouple the beamforming vector and the position coefficient of PA. For the given beamforming vectors, a new objective function is formulated with respect to the positions of the PAs. With the characteristics of the formulated objective function, the optimization problem for the position coefficients of PAs can be decomposed into multiple independent subproblems, each corresponding to a PA's position coefficient, and a projected gradient descent (PGD)-based method, constrained by the feasible movable region of each PA, is then developed to obtain the suboptimal position coefficients. The performance improvements achieved by the combination of PAS and SLP, as well as the effectiveness of the proposed algorithm are verified through the simulation results.

Terahertz (THz) communication can offer terabit-per-second rates in future wireless systems, thanks to the ultra-wide bandwidths, but require large antenna arrays. As antenna apertures expand and we enter the near-field scenarios, the conventional binary classification of communication links as either Line-of-Sight (LoS) or Non-Line-of-Sight (NLoS) becomes insufficient. Instead, quasi-LoS scenarios, where the LoS path is partially obstructed, are increasingly prevalent, posing significant challenges for traditional LoS focusing and steering beams. The Airy beam serves as a promising alternative, utilizing its non-diffracting and curved trajectory properties to mitigate such blockages. However, while existing electromagnetics literature primarily explores their physical patterns without practical generation schemes, recent communication-oriented designs predominantly rely on learning-based frameworks lacking interpretable closed-form solutions. To address this issue, this paper investigates a closed-form Airy beam design to efficiently synthesize Airy beam phase profiles based on the positions of the transceivers and obstacles. Specifically, rigorous analytical derivations of the electric field and trajectory are presented to establish a deterministic closed-form design for ULA Airy beamforming. Leveraging 3D wavefront separability, this framework is extended to uniform planar arrays (UPAs) with two operation modes: the hybrid focusing-Airy mode and the dual Airy mode. Simulation results verify the effectiveness of our derived trajectory equations and demonstrate that the proposed closed-form design significantly outperforms conventional beamforming schemes in quasi-LoS scenarios. Furthermore, the proposed method achieves performance comparable to exhaustive numerical searches with low computational complexity and enhanced physical interpretability.

Beyond-diagonal reconfigurable intelligent surfaces (BD-RISs) are an emerging RIS 2.0 technology for future wireless communication. However, BD-RISs are primarily passive without active amplification, suffering from severe multiplicative path loss. To address the concern of multiplicative path loss, in this work we investigate the active BD-RIS including the modeling, architecture design, and optimization. We first analyze the active BD-RIS using multiport network theory with scattering parameters and derive a physical and electromagnetic compliant active BD-RIS aided communication model. We also design two new active BD-RIS architectures, namely fully- and group-connected active BD-RISs. Based on the proposed model and architecture, we investigate the active BD-RIS aided single-input single-output system and derive the closed-form optimal solution and scaling law of the signal-to-noise ratio. We further investigate the active BD-RIS aided multiple-input multiple-output system and propose an iterative algorithm based on quadratically constrained quadratic programming to maximize the spectral efficiency. Numerical results are provided and show that the active BD-RIS can achieve higher spectral efficiency than the active/passive diagonal RIS and passive BD-RIS. For example, to achieve the same spectral efficiency, the number of elements required by active BD-RIS is less than half of that required by active diagonal RIS, showing the advantages of active BD-RIS.

This study provides a comprehensive synthesis of Artificial Intelligence (AI), especially Machine Learning (ML) and Deep Learning (DL), in ransomware defense. Using a "review of reviews" methodology based on PRISMA, this paper gathers insights on how AI is transforming ransomware detection, prevention, and mitigation strategies during the past five years (2020-2024). The findings highlight the effectiveness of hybrid models that combine multiple analysis techniques such as code inspection (static analysis) and behavior monitoring during execution (dynamic analysis). The study also explores anomaly detection and early warning mechanisms before encryption to address the increasing complexity of ransomware. In addition, it examines key challenges in ransomware defense, including techniques designed to deceive AI-driven detection systems and the lack of strong and diverse datasets. The results highlight the role of AI in early detection and real-time response systems, improving scalability and resilience. Using a systematic review-of-reviews approach, this study consolidates insights from multiple review articles, identifies effective AI models, and bridges theory with practice to support collaboration among academia, industry, and policymakers. Future research directions and practical recommendations for cybersecurity practitioners are also discussed. Finally, this paper proposes a roadmap for advancing AI-driven countermeasures to protect critical systems and infrastructures against evolving ransomware threats.

This paper investigates joint trajectory and active beamforming design for unmanned aerial vehicle (UAV)-enabled ultra-reliable low-latency communication (URLLC) systems under finite blocklength (FBL) transmission. Unlike conventional Shannon-capacity formulations, the FBL regime introduces a signal-to-interference-plus-noise ratio (SINR)-dependent dispersion penalty that increases the sensitivity of reliability to mobility-induced channel variations. To address this challenge, we develop a propulsion-aware model predictive control (MPC) framework that performs receding-horizon joint trajectory and multi-user beamforming optimization while enforcing FBL-based rate constraints. The resulting long-horizon nonconvex problem is decomposed into beamforming and trajectory subproblems using alternating optimization. Concave surrogate is constructed for the Shannon-capacity term, while convex approximations are derived for the dispersion term and the nonlinear propulsion power model, yielding tractable convex subproblems solved iteratively. Compared with an offline MPC baseline, where the predictive problem is solved once over the entire mission horizon without feedback updates, and a conventional offline trajectory-beamforming optimization, the proposed closed-loop framework achieves disturbance-resilient mission completion under UAV position disturbances. Simulation results show that, compared with maximum ratio transmission (MRT) and equal-power allocation, the proposed interference-aware design significantly improves URLLC reliability under stringent minimum rate constraints. The results also quantify the impact of antenna scaling, transmit power, and transmission time on FBL performance, providing insights for reliability-centric UAV-enabled wireless networks in 5G and beyond.

Physical layer key generation (PLKG) has emerged as a promising solution for achieving highly secured and low-latency key distribution, offering information-theoretic security that is inherently resilient to quantum attacks. However, simultaneously ensuring a high data transmission rate and a high secret key generation rate under eavesdropping attacks remains a major challenge. In time-division duplex (TDD) systems with multiple antennas, we derive closed-form expressions for both rates by modeling the legitimate channel as a time-correlated autoregressive (AR) process. This formulation leads to a highly nonconvex and time-coupled optimization problem, rendering traditional optimization methods ineffective. To address this issue, we propose a multi-agent soft actor-critic (SAC) framework equipped with a long short-term memory (LSTM) adversary prediction module to cope with the partial observability of the eavesdropper's mode. Simulation results demonstrate that the proposed approach achieves superior performance compared with other benchmark algorithms, while effectively balancing the trade-off between secret key generation rate and data transmission rate. The results also confirm the robustness of the proposed framework against intelligent eavesdropping and partial observation uncertainty.

Wave-based signal processing conventionally encodes input data into the input wavefront, making it challenging to implement non-linear operations. Programmable wave systems enable an alternative approach: encoding the input data into the scattering properties of tunable components. With such structural input encoding, two potentially non-linear mappings are involved: first, from the input data to the tunable components' scattering characteristics, and, second, from these scattering characteristics to the output wavefront. In this paper, we systematically examine the expressivity of a wave-based physical neural network (WPNN) with structural input encoding. Our analysis is based on a physics-consistent multiport-network model of a compact D-band rich-scattering cavity parametrized by a 100-element programmable metasurface. We separately control encoding non-linearity, structural non-linearity, and network depth in order to examine their interplay, considering a controlled scalar regression task. With phase encoding and strong inter-element mutual coupling (MC), both aforementioned mappings are strongly non-linear and the WPNN performs very well even with a single layer. We further observe that additional layers can partially compensate for weak inter-element MC. In addition, we demonstrate that WPNN depth can improve expressivity even when it is not associated with an increase in trainable weights. Altogether, our results provide a physics-consistent picture of how encoding choice, MC strength, and depth jointly govern the expressive power of PM-based WPNNs, informing design choices for future experimental implementations of WPNNs.

This paper develops a comprehensive target modeling, beamforming optimization, and parameter estimation framework for extended-target sensing in wideband MIMO-OFDM integrated sensing and communication systems. We propose a parametric scattering model (PSM) that decouples target geometry from electromagnetic scattering characteristics, requiring only six nonlinear geometric parameters and linear radar cross-section (RCS) coefficients. Based on this compact structure, we derive a hybrid Bayesian Cramér-Rao bound (CRB) for joint estimation of azimuth, elevation, and range-related parameters. To handle inherent range ambiguities due to OFDM signaling, we analyze the range ambiguity function and introduce range sidelobe suppression constraints around the true range. Based on these constraints, we formulate an ambiguity-aware transmit beamforming design that minimizes a weighted geometric CRB subject to per-user signal-to-interference-plus-noise ratio (SINR) requirements and a total power budget. As benchmarks, we extend two other common models to the same wideband MIMO-OFDM scenario. We also derive maximum a posteriori estimators and a computational complexity analysis for all three models. Simulation results demonstrate that the proposed PSM-based approach achieves improved target localization with significantly reduced runtime for beamforming optimization and parameter estimation, while consistently satisfying communication SINR requirements.

Authenticity and condition assessment are central to conservation decision-making, yet interpretation and reporting of thermographic output remain largely bespoke and expert-dependent, complicating comparison across collections and limiting systematic integration into conservation documentation. Pulsed Active Infrared Thermography (AIRT) is sensitive to subsurface features such as material heterogeneity, voids, and past interventions; however, its broader adoption is constrained by artifact misinterpretation, inter-laboratory variability, and the absence of standardized, explainable reporting frameworks. Although multi-modal thermographic processing techniques are established, their integration with structured natural-language interpretation has not been explored in cultural heritage. A fully automated thermography-vision-language model (VLM) framework is presented. It combines multi-modal AIRT analysis with modality-aware textual reporting, without human intervention during inference. Thermal sequences are processed using Principal Component Thermography (PCT), Thermographic Signal Reconstruction (TSR), and Pulsed Phase Thermography (PPT), and the resulting anomaly masks are fused into a consensus segmentation that emphasizes regions supported by multiple thermal indicators while mitigating boundary artifacts. The fused evidence is provided to a VLM, which generates structured reports describing the location of the anomaly, thermal behavior, and plausible physical interpretations while explicitly acknowledging the uncertainty and diagnostic limitations. Evaluation on two marquetries demonstrates consistent anomaly detection and stable structured interpretations, indicating reproducibility and generalizability across samples.

In this paper, we propose a joint delay-Doppler estimation framework for Rydberg atomic quantum receivers (RAQRs) leveraging affine frequency division multiplexing (AFDM), as a future enabler of hyper integrated sensing and communication (ISAC) in 6G and beyond. The proposed approach preserves the extreme sensitivity of RAQRs, while offering a pioneering solution to the joint estimation of delay-Doppler parameters of mobile targets, which has yet to be addressed in the literature due to the inherent coupling of time-frequency parameters in the optical readout of RAQRs to the best of our knowledge. To overcome this unavoidable ambiguity, we propose a dual-chirp AFDM framework where the utilization of distinct chirp parameters effectively converts the otherwise ambiguous estimation problem into a full-rank system, enabling unique delay-Doppler parameter extraction from RAQRs. Numerical simulations verify that the proposed dual-chirp AFDM shows superior delay-Doppler estimation performance compared to the classical single-chirp AFDM over RAQRs.

A scattering event in a quantum field theory is a coherent superposition of all processes consistent with its symmetries and kinematics. While real-time simulations have progressed toward resolving individual channels, existing approaches rely on knowledge of the asymptotic particle wavefunctions. This work introduces an experimentally inspired method to isolate scattering channels in Matrix Product State simulations based on the entanglement structure of the late-time wavefunction. Schmidt decompositions at spatial bipartitions of the post-scattering state identify elastic and inelastic contributions, enabling deterministic detection of outgoing particles of specific species. This method may be used in settings beyond scattering and is applied to detect heavy particles produced in a collision in the one-dimensional Ising field theory. Natural extensions to quantum simulations of other systems and higher-order processes are discussed.

Code-switching offers a route to universal, fault-tolerant quantum computation by circumventing the limitation implied by the Eastin-Knill theorem against a universal transversal gate set within a single quantum code. Here, we present a fault-tolerant code-switching protocol between two versions of the $[[8, 3, 2]]$ code. One version supports weakly fault-tolerant single-qubit Clifford gates, while the other supports a logical $\overline{\mathrm{CCZ}}$ gate via transversal $T/T^\dagger$ together with logical $\overline{\mathrm{CZ}}$, $\overline{\mathrm{CNOT}}$, and $\overline{\mathrm{SWAP}}$ gates. Because both codes have distance 2, the protocol operates in a postselected, error-detecting regime: single faults lead to detectable outcomes, and accepted runs exhibit quadratic suppression of logical error rates. This yields a universal scheme for postselected fault-tolerant computation. We validate the protocol numerically through simulations of state preparation, code switching, and a three-logical-qubit implementation of Grover's search.

A central goal of quantum computation is the realistic simulation of quantum materials. Although quantum processors have advanced rapidly in scale and fidelity, it has remained unclear whether pre-fault-tolerant devices can perform quantitatively reliable material simulations within their limited gate budgets. Here, we demonstrate that a superconducting quantum processor operating on up to 50 qubits can already produce meaningful, quantitative comparisons with inelastic neutron-scattering measurements of KCuF$_3$, a canonical realization of a gapless Luttinger liquid system with a strongly correlated ground state and a spectrum of emergent spinons. The quantum simulation is enabled by a quantum-classical workflow for computing dynamical structure factors (DSFs). The resulting spectra are benchmarked against experimental measurements using multiple metrics, highlighting the impact of circuit depth and circuit fidelity on simulation accuracy. Finally, we extend our simulations to 1D XXZ Heisenberg model with next-nearest neighbor interactions and a strong anisotropy, producing a gapped excitation spectrum, which could be used to describe the CsCoX$_3$ compounds above the Néel temperature. Our results establish a framework for computing DSFs for quantum materials in classically challenging regimes of strong entanglement and long-range interactions, enabling quantum simulations that are directly testable against laboratory measurements.

Bouncing geodesics have been used as valuable probes of black hole singularities. In the dual boundary theory, the presence of bouncing geodesics is encoded in the analytic structure of correlation functions. Thus, when their existence is related to the presence of a black hole singularity, this presents a practical holographic framework to analyse, diagnose, and classify spacetimes with curvature singularities. To make this intuition precise, we use the Hadamard theory of hyperbolic differential equations to prove that both bulk and boundary retarded propagators diverge whenever two points can be connected by a null geodesic. We clarify why this statement remains valid beyond the geodesic regime (for operators of any dimension) and examine how holographic renormalisation modifies the structure of the dual propagator. We also present a general characterisation of bouncing geodesics and the associated singularities in correlation functions for arbitrary spacetimes. Furthermore, we compare the analytic structure of the correlators in position and momentum space and discuss explicit examples. Finally, we demonstrate the validity and concrete limitations of the bouncing geodesic approach to the study of black hole singularities. In particular, we show an explicit example of a black hole in the self-dual linear axion model, which has a curvature singularity despite the absence of bouncing geodesics.

Useful quantum information processing ultimately requires operations over large Hilbert spaces, where logical information can be encoded efficiently and protected against noise. Harmonic oscillators naturally provide access to such high-dimensional spaces and enable hardware-efficient, error-correctable bosonic encodings. However, direct entangling operations between oscillators remains an outstanding challenge. Existing strategies typically rely on parametrically activating interactions that populate the excited states of an ancillary nonlinear element. This induces an effective interaction between the oscillators, at the expense of introducing additional dissipation channels and potential leakage from the encoded manifold. Here, we engineer a Raman-assisted cross-Kerr interaction between microwave photons hosted in two superconducting cavities, without exciting the nonlinear element, thereby suppressing coupler-induced decoherence.This approach generates a direct coupling between microwave photons that is exploited to implement a controlled-phase gate within the single- and two-photon subspaces of two oscillators, directly entangling them. Finally, we harness this dynamics to map the photon-number parity of a storage cavity onto an auxiliary oscillator rather than a nonlinear element, enabling error detection while protecting the storage mode from measurement-induced decoherence. Our work expands the bosonic circuit quantum electrodynamics (cQED) toolbox by enabling coherence-preserving direct photon-photon interactions between oscillators. This realizes an entangling gate that operates entirely within a bosonic code space while suppressing decoherence from nonlinear ancilla excitations, providing a key primitive for fault-tolerant bosonic quantum computing.

We propose a Monte Carlo based method for simulating the open system dynamics of multiple exchange-only (EO) qubits. In the EO encoding, the total spin projection quantum number along the $z$-axis of the three constituent spins remains unchanged under exchange operations, in contrast to the open system (or multi-qubit miscalibration) setting where coherent and incoherent mixing of states with different quantum numbers occurs. In our approach, we choose to measure the total spin component along the $z$-axis of each EO qubit after every logical quantum operation, which decoheres coherent mixtures of states with different spin projection quantum numbers. Independent simulations thus give different trajectories of the system in the associated subspaces, so we refer to this method as the Subspace Monte Carlo method. With each EO qubit having a definite spin projection quantum number, the density matrix of $n$ qubits can be represented by a vector of dimension $3^{2n}$, instead of $8^{2n}$, with an additional vector of dimension $n$ to label the quantum number of each qubit. We show that this approximation of the dynamics remains faithful to the true dynamics when the simulated circuits twirl the noise, converting coherent errors to stochastic errors, which can be achieved using randomized compiling. We use this simulation approach to study how correlations in measurement outcomes of circuits with reset-if-leaked gadgets, such as a multi-round Bell state stabilization circuit that uses 6 EO qubits, are affected by the choice of CNOT implementations.

Realizing error-corrected logical qubits is a central goal for the current development of digital quantum computers. Neutral atoms offer the opportunity to coherently shuttle atoms for realizing efficient quantum error correction based on long-range connectivity and parallel atom transport. Nevertheless, time overheads in shuttling atoms and complex control hardware pose challenges to scaling current architectures. Here, we introduce atom velocity as a new degree of freedom in neutral-atom architectures tailored to quantum error correction. Through controlled Doppler shifts, we demonstrate velocity-selective mid-circuit state preparation and measurement on moving atoms, leaving spectator atoms unaffected. Furthermore, we achieve on-the-fly local single-qubit rotations by mapping micron-scale atom displacements to the spatial phase of global control beams. Complementing these techniques with CZ entangling gates with a fidelity of 99.86(4)%, we experimentally implement key primitives for quantum error correction and measurement-based quantum computing. We generate an eight-qubit entangled cluster state with an average stabilizer value of 0.830(4), realize an [[4,2,2]] error-detection code with 99.0(3) % logical Bell-state fidelity, and perform stabilizer measurements using a flying ancilla. By enabling selective operations on continuously moving atoms using only global beams, this velocity-enabled architecture reduces hardware overhead while minimizing shuttling and transfer delays, opening a new pathway for fast, large-scale atom-based quantum computation.

In resetting dynamics, a system is repeatedly coupled to and decoupled from ancillary degrees of freedom that are reinitialized between interactions. This provides a versatile route to engineer nonequilibrium steady states and constitutes a powerful and analytically transparent framework for studying nonequilibrium dynamics in quadratic fermionic models. The baseline noninteracting resetting scheme yields an affine evolution for the subsystem single-particle density matrix (SPDM), with a clear operational interpretation: a finite environment block E mediates the interaction between the subsystem S and an ideal external thermal reservoir. In this work, we develop a controlled extension of such a framework to weakly interacting systems. We introduce a Hartree mean-field treatment of density-density interactions that preserves closure of the SPDM dynamics while producing genuinely nonlinear behavior. We further construct a completely positive (CP-safe) Gaussian Lindblad embedding that reproduces the resetting dynamics in the noninteracting limit and yields a continuous-time representation of environmental thermalization when interactions are present. Our analytical results are complemented by numerical studies of a ring segmentation geometry and a minimal two-site model, revealing interaction-enabled steady states that cannot be obtained in any purely quadratic setting. Together, these results establish a general and physically consistent route for incorporating weak interactions ino resetting-based approaches to open quantum system.

We develop a framework that enables direct and meaningful comparison of two early fault-tolerant methods for the computation of eigenenergies, namely \gls{qksd} and \gls{spe}, within which both methods use expectation values of Chebyshev polynomials of the Hamiltonian as input. For \gls{qksd} we propose methods for optimally distributing shots and ensuring sufficient non-linearity of states spanning the Krylov space. For \gls{spe} we improve rigorous error-bounds, achieving roughly a factor $2/3$ reduction of circuit depth. We provide insights into the scalability of and the practical realization of these methods by computing the maximum Chebyshev degree, linearly related to circuit depth, and the respective number of repetitions required for the simulation of molecules with active spaces up to 54 electrons in 36 orbitals by leveraging \gls{mps}/\gls{dmrg}.

In this work we investigate discrete structures in product Hilbert spaces. For monopartite systems of size $d$ one relies on the Weyl-Heisenberg group $WH(d)$, while in the case of composite Hilbert spaces we identify designs covariant with respect to the product group, $[WH(p)]^{\otimes n}$. In analogy with magic -a quantity attaining its maximum for states fiducial with respect to $WH(d)$ -we introduce a similar notion of magick, defined with respect to the product group. The maximum of this quantity over all equimodular vectors yields fiducial states that generate $d$ $\textit{a priori}$ isoentangled mutually unbiased bases (MUBs), which, when supplemented by the identity, form their complete set. Such fiducial states are explicitly constructed in all prime-power dimensions $p^n$ with $p\ge 3$. The result for $p\ge 5$ extends the construction of Klappenecker and Rötteler, whereas for $p=3$ it is mathematically distinct and is based on Galois rings. The global maximum of magick for $d=2^3$ yields fiducial states corresponding to the symmetric informationally complete (SIC) generalized measurement of Hoggar. Our approach feeds into a unifying perspective in which highly symmetric quantum designs emerge from fiducial states with extremal properties via structured group-orbit constructions.

We introduce QiboAgent, a reference implementation designed to serve as a practitioner's guideline for developing specialized coding assistants in Quantum Computing middleware. Addressing the limitations in scientific software development of general-purpose proprietary models, we explore how lightweight, open-source Large Language Models (LLMs) provided with a custom workflow architecture compare. In detail, we experiment with two complementary paradigms: a Retrieval-Augmented Generation pipeline for high-precision information retrieval, and an autonomous agentic workflow for complex software engineering tasks. We observe that this hybrid approach significantly reduces hallucination rates in code generation compared to a proprietary baseline, achieving a peak accuracy of 90.2% with relatively small open-source models of size up to 30B parameters. Furthermore, the agentic framework exhibits advanced coding capabilities, automating the resolution of maintenance issues and new features requests, or by prototyping larger-scale refactors of the codebase, such as producing a compiled Rust module with bindings of an original pure python package, Qibo in our case. The LLM workflows used for our analysis are integrated into a user interface and a Model Context Protocol server, providing an accessible tool for Qibo developers.

Quantum annealing processors typically control qubits in unison, attenuating quantum fluctuations uniformly until the applied system Hamiltonian is diagonal in the computational basis. This simplifies control requirements, allowing annealing QPUs to scale to much larger sizes than gate-based systems, but constraining the class of available operations. Here we expand the class by performing analog-digital quantum computing in a highly-multiplexed, superconducting quantum annealing processor. This involves evolution under a fixed many-body Hamiltonian that, in the weak-coupling regime, is well-described by an effective XY model, together with arbitrary-basis initialization and measurement via auxiliary qubits. Operationally, this is equivalent to implementing single-qubit gates at the beginning and end of an analog quantum evolution. We demonstrate this capability with several foundational applications: single-qubit and two-qubit coherent oscillations with varying initialization and measurement bases, a multi-qubit quantum walk with fermionic dispersion in line with theory, and Anderson localization in a disordered chain. These experiments open the door to a wide range of new possibilities in quantum computation and simulation, greatly expanding the applications of commercially available quantum annealing processors.

Multispectral satellite imagery poses significant challenges for deep learning models due to the high dimensionality of spectral data and the presence of structured correlations across channels. Recent work in quantum machine learning suggests that unitary evolutions and Hilbert-space embeddings can introduce useful inductive biases for learning. In this work, we show that several empirical advantages often attributed to quantum feature maps can be more precisely understood as consequences of geometric structure induced by unitary group actions and the associated quotient symmetries. Motivated by this observation, we introduce a fully classical pooling mechanism that maps latent features to complex projective space via a fixed-reference unitary action. This construction effectively collapses non-identifiable degrees of freedom, leading to a reduction in the dimensionality of the learned representations. Empirical results on multispectral satellite imagery show that incorporating this quantum-inspired pooling operation into a convolutional neural network improves optimization stability, accelerates convergence, and substantially reduces variance compared to standard pooling baselines. These results clarify the role of geometric structure in quantum-inspired architectures and demonstrate that their benefits can be reproduced through principled geometric inductive biases implemented entirely within classical deep learning models.

Recent advances in optical imaging and communication increasingly involve high-dimensional, partially coherent light, creating a growing need for scalable tools to measure and manipulate coherence. Here, we demonstrate the automatic separation of spatially partially coherent light into "coherence modes" -- its orthogonal and mutually incoherent components. To make this separation possible, we exploit variational processing in layered self-configuring interferometer architectures in a silicon photonic circuit. This process formally finds and measures the eigenvectors and eigenvalues of the coherency matrix, hence measuring the partially coherent state, while leaving it intact and separated after optimization. Furthermore, we show that mutually incoherent beams, if spatially orthogonal, can be automatically separated even if they are completely overlapped, hence separating unknown laser beams based only on their mutual incoherence. Our experiment finds and separates the two strongest coherence modes starting from a nine-mode sampling of the partially or fully overlapping fields from two independent lasers. The method requires a number of physical components that scales linearly with the rank $r$ of the coherency matrix and operates through a sequence of $r$ in situ gradient-based optimizations enabled by electronic drive frequency multiplexing of interferometer phase shifters. We benchmark its performance against a mixture-based tomographic method, also implemented on chip. These results establish a scalable framework for programmable coherence analysis and control in imaging, communication, and photonic information processing.

Solving large-scale sparse linear systems is a challenging computational task due to the introduction of non-zero elements, or "fill-in." The Graph Partitioning Problem (GPP) arises naturally when minimizing fill-in and accelerating solvers. In this paper, we measure the end-to-end performance of a hybrid quantum-classical framework designed to accelerate Finite Element Analysis (FEA) by integrating a quantum solver for GPP into Synopsys/Ansys' LS-DYNA multiphysics simulation software. The quantum solver we use is based on Iterative-QAOA, a scalable, non-variational quantum approach for optimization. We focus on two specific classes of FEA problems, namely vibrational (eigenmode) analysis and transient simulation. We report numerical simulations on up to 150 qubits done on NVIDIA's CUDA-Q/cuTensorNet and implementation on IonQ's Forte quantum hardware. The potential impact on LS-DYNA workflows is quantified by measuring the wall-clock time-to-solution for complex problem instances, including vibrational analysis of large finite element models of a sedan car and a Rolls-Royce jet engine, as well as transient simulations of a drill and an impeller. We performed end-to-end performance measurements on meshes comprising up to 35 million elements. Measurements were conducted using LS-DYNA in distributed-memory mode via Message Passing Interface (MPI) on AWS and Synopsys compute clusters. Our findings indicate that with a quantum computer in the loop, amortized LS-DYNA wall-clock time can be improved by up to 15% for specific cases and by at least 7% for all models considered. These results highlight the significant potential of quantum computing to reduce time-to-solution for large-scale FEA simulations within the Noisy Intermediate-Scale Quantum (NISQ) era, offering an approach that is scalable and extendable into the fault-tolerant quantum computing regime.

Stochastic gravitational-wave backgrounds (SGWBs) of primordial origin offer a powerful probe of early-Universe physics and possible dark-sector dynamics. While most searches focus on the GW power spectrum, additional information is encoded in higher-order correlators that characterize the statistical properties of the signal. In this work we study non-Gaussian features of a cosmological SGWB generated at second order by vector fluctuations, a class of sources well motivated in early-Universe scenarios. Within this framework we develop tools to characterize higher-order GW correlators and compute representative four-point functions that generate a connected contribution to the GW trispectrum. We show that the trispectrum amplitude scales as the square of the GW power spectrum and peaks in characteristic folded momentum configurations, reflecting the structure of the nonlinear source. We then explore the observational implications. First, we demonstrate that the connected trispectrum contributes to the variance of two-point overlap reduction functions, including the Hellings-Downs curve relevant for pulsar timing arrays. We then construct the optimal estimator to measure the connected trispectrum with ground-based interferometers. Our results highlight how non-Gaussian SGWB statistics provide a complementary observable to probe the origin of GW backgrounds and to distinguish cosmological from astrophysical sources.

Simplifying composite open quantum systems through model reduction is central to enable their analytical and numerical understanding. In this work, we introduce a self-consistent approach to eliminate the cavity degrees of freedom of cavity quantum electrodynamics (CQED) devices in the non-adiabatic regime, where the cavity memory time is comparable with the timescales of the atom dynamics. To do so, we consider a CQED system consisting of a two-level atom coupled to a single-mode cavity, both subsystems interacting with the environment through an arbitrary number of ports, within the input-output formalism. A self-consistency equation is derived for the reduced atom dynamics. This allows retrieving an exact expression for the effective Purcell-enhanced emission rate and, under reasonable approximations, a set of self-consistent dynamical equations and input-output relations for the effective two level atom. The resulting reduced model captures non-Markovian features, characterized through an effective Lindblad equation exhibiting two decoherence rates, a positive and a negative one. In the continuous-wave excitation regime, we benchmark our approach by computing effective steady states and output flux expressions beyond the low-power excitation regime, for which a semi-classical treatment is usually applied. We also compute two-time correlations and spectral densities, showing an excellent agreement with full cavity quantum electrodynamics simulations, except in the strong-coupling, high-excitation regime. Our results provide a practical framework for reducing the size of CQED models, which could be generalized to more complex atom and cavity configurations.

We present a general framework for promoting first-order pulse sequences in quantum simulation to higher-order sequences that maintain robustness in the presence of finite pulse-width effects. Our approach maps a given first-order pulse sequence to a first-order Trotter formula, applies higher-order Trotter-formula constructions, and then compiles the resulting evolution back into physically implementable finite-width pulses via dynamically corrected gates. The resulting sequences achieve arbitrarily high-order error scaling with respect to the control cycle time of the underlying first-order sequence while maintaining robustness to finite pulse-width effects. The framework also enables the use of multi-product formulas for more efficient constructions. We apply the framework to several physically motivated quantum-simulation tasks and numerically verify the predicted error scalings.

We present a systematic workflow for the design of one-dimensional photonic crystal nanobeam cavities with non-zero cavity lengths. By simultaneously optimizing the lattice periodicity, air-hole geometry, and cavity length, our approach enables precise control of optical confinement while mitigating radiative losses and linewidth broadening effects. The method is further extended to the design of crossed nanobeam cavities with both matching and mismatched resonance frequencies. This strategy significantly reduces the need for extensive parameter sweeps, providing an efficient route toward optimized cavity designs for integrated quantum photonic applications. Moreover, the resulting structures are inherently compatible with the integration of single-photon emitters.

Preparing low energy states is a central challenge in quantum computing and quantum complexity theory. Several known approaches to prepare low energy states often get stuck in suboptimal states, such as high energy eigenstates (or low variance high energy states). We develop a heuristic method to go past this barrier for local Hamiltonian systems with relatively low frustration, by taking advantage of the fact that such systems come with multiple Hamiltonians that agree on the low-energy subspaces. We establish an energy-based uncertainty principle, which shows that these Hamiltonians in fact do not have common eigenstates in the high energy regime. This allows us to run energy lowering steps in an alternating manner over the Hamiltonians. We run numerical simulations to check the performance of the `alternating' algorithm on small system sizes, for the 1D AKLT model and instances of Heisenberg model on general graphs. We also formulate a version of the energy-based uncertainty principle using sparse Hamiltonians, which shows a quadratically larger variance at higher energies and hence leads to a larger energy change. We use this version to simulate the method on energy profiles with high energy barriers.

We investigate the non-adiabatic quantum dynamics of ultralong-range Rydberg molecules using a vibronically coupled two-channel treatment. The two channels are composed of coupled trilobite and butterfly electronic states, formed as a result of $S$-wave and $P$-wave scattering of high angular momentum Rydberg electrons with perturbing ground state atoms. Within the Born-Oppenheimer treatment, the $P$-wave scattering channel introduces an adiabatic decay pathway that affects the stability and lifetimes of trilobite states. Our numerical results show that the vibronic coupling is dependent on the principal quantum number $n$, and for certain $n$ there is non-adiabatic stabilization against internal molecular decay, facilitating previously studied dynamical effects in pure trilobite molecules. Apart from the internal diffraction effect we also observe interesting multi-well tunneling effects, during low-energy oscillations for certain $n$-values. Our work serves to highlight that the unique $R$-dependent electronic structure of these polar molecules, along with high level densities, promise many exciting dynamical effects.

We present a quantum field theoretic formulation of acoustoelectric interactions in waveguide-like systems of arbitrary cross-section. Building on an open quantum systems approach, we derive a unified description of plasmon-phonon coupling that incorporates dissipation, noise, and the influence of drift currents. Our analysis captures both bulk and surface plasmon modes, highlighting how drift currents Doppler-shift plasmonic resonances and reshape the phonon noise spectrum. The resulting Heisenberg-Langevin equations yield closed-form expressions for frequency shifts, gain, and noise power spectra, enabling direct evaluation of performance metrics such as the noise factor in acoustoelectric amplifiers and oscillators. In the appropriate limits, this framework reproduces known results while extending them to complex geometries.

The merger-ringdown radiation from a black hole binary merger is accurately modeled by a sum of linear quasinormal modes (QNMs). Recently, a non-QNM ``direct wave" component of the radiation, associated with prompt emission from a plunging perturber, has been identified. Motivated by the behavior of null geodesics perturbed from the remnant light ring, the Backwards One Body (BOB) approach has been shown to model the full merger-ringdown radiation to high accuracy, while using only a minimal number of parameters. In this work, using the Pöschl--Teller potential, we first show how the BOB amplitude evolution can be recovered from the QNM pole contributions. We then apply rational filters to isolate the non-QNM content in BOB and numerical relativity waveforms. We show that BOB naturally captures the direct wave component of the merger radiation, explaining its accuracy near the waveform peak. Finally, we use BOB to show that the direct wave frequency is largely uncorrelated with the horizon frequency, even for high spin remnants, and instead tracks the News frequency at the time of the peak News amplitude.

Compact, customizable, non-magnetic vacuum systems are a key requirement for many field applications of quantum technology based on cold atoms. We report on the development and construction of a compact, low-cost ultra-high vacuum compatible cell using the glass-ceramic MACOR. The cell offers a CF flange connection to commercial vacuum technology, as well as high numerical aperture viewports for precision optical measurements. The presented technology shows stable vacuum pressures of $< 1 \cdot 10^{-10}$ mbar for more than a year since the implementation into the vacuum system of a quantum gas experiment, further proving suitability for general quantum technology applications.

The large-distance development of a sandwich gravitational wave, consistent with Carroll symmetry, provides us with a surprisingly good analytic approximation of the motion of particles in a wave with Pöschl-Teller profile. The role of the 2nd solution of the Stern-Liouville equation is highlighted. Similar results hold for Gaussian profiles.

The Einstein Telescope (ET) is a third-generation underground gravitational-wave observatory designed to extend the detection sensitivity down to a few Hertz. Newtonian noise is expected to limit the low-frequency sensitivity of ET, particularly in the 1.7-6 Hz band. Most existing estimates rely on analytical or semi-analytical models assuming homogeneous or layered media, neglecting geological heterogeneity and complex wave interactions. In this work, we present a numerical framework for Newtonian-noise estimation based on spectral-element simulations of a seismic wave field. As a proof of concept, we first benchmark the numerical results against analytical plane-wave predictions in a two-dimensional homogeneous medium with a single surface source, demonstrating excellent agreement for both bulk and cavern contributions. We then extend the model to an array of 30 stochastic surface sources to approximate stationary ambient seismic excitation. The P-wave fraction inferred from the simulated wave field is, in this simple homogeneous case, significantly lower than commonly assumed, indicating enhanced prospects for Newtonian-noise mitigation. The framework is readily applicable to three-dimensional simulations and to integration of detailed local seismic models and topography, offering strong potential for site-specific Newtonian-noise estimation.

A puzzling implication of current data interpreted in the $Λ$CDM cosmology is the preference for a negative sum of neutrino masses. Moving to $w_0w_a$CDM brings an appreciable fraction of the neutrino mass posterior back to positive values, while the constant equation-of-state dark energy case $w$CDM does not. We investigate a variety of one-parameter dark energy equations of state (DE EoS), each variation with particular physical properties, to understand whether a two-parameter DE EoS is required to bring the neutrino mass positive. The conclusion is that certain one-parameter DE EoS can suffice, implying that the data are pointing toward physical characteristics rather than a broad degeneracy. The required characteristics are identified as phantom dark energy at high redshift, crossing $w=-1$ at lower redshift.

We develop a Bohmian analysis of a two-dimensional ghost Hamiltonian and its mapping to the degenerate Pais-Uhlenbeck model. Using Gaussian wavepackets, we derive the corresponding guidance equations, the centre and width evolution, and the quantum potential. We use these quantities to characterise bounded, quasi-semiclassical, spiral, and runaway regimes. The Bohmian trajectories provide a direct dynamical diagnostic of coherence, packet deformation, and quantum-classical separation. We then compare a bi-Hamiltonian pair consisting of the ghost Hamiltonian and a classically equivalent alternative formulation. While the two descriptions produce identical classical trajectories, they lead to different Bohmian trajectories and different quantum potentials evaluated along those trajectories. This demonstrates that classical equivalence need not extend to Bohmian quantum dynamics and identifies a concrete quantum ambiguity in the degenerate higher-derivative system.

The SU(2) Wigner function represents a quantum state of a spin-$J$ as a real-valued function on the surface of a 2-sphere. For an ensemble of $N$ spin-1/2 particles, this representation is useful when the dynamics is restricted to a single SU(2) irrep, e.g., the symmetric subspace with $J=N/2$. Physically relevant noise sources tend to be local, such as spontaneous emission, depolarizing, and incoherent optical pumping, all of which transfer the state outside of the initial irrep, and as such the SU(2) Wigner function is no longer a useful representation. In this work, we address this issue by encoding a noisy spin ensemble in an SU(3) irrep, and evaluating the SU(3) Wigner function for that irrep. We find that physical constraints enforced by the noise eliminate all but three real parameters from the input to the Wigner function, which can then be interpreted as a polar, azimuthal, and radial component. This interpretation leads us to refer to the resulting Wigner function as the solid spin Wigner function, visualized on a solid ball rather than a hollow sphere.

We explain why Calabi-Yau and complete elliptic integrals do not contribute to conservative observables at fifth post-Minkowskian order, despite appearing in intermediate steps. At even loop orders, conservative contributions are tied to terms proportional to the logarithm of the momentum transfer, which in dimensional regularization arise from singular regions. We show that in the classical limit, the integral classes responsible for Calabi-Yau and complete elliptic behavior are absent from the ultraviolet singular structures that generate the required logarithm. This perspective also suggests alternative strategies for analyzing the classical limit of multiloop integrals in the conservative sector at even loop orders.

Dipolar molecules in optical traps are a versatile platform for studying many-body phases of quantum matter in the presence of strong and long-range interactions. The dipolar interactions in such setups can be enabled by microwave driving opposite parity rotational levels of the molecules. We find that the regime where the $N=0,J=1/2,F=1$ state is coupled to the $N=1,J=3/2,F=2$ manifold with circularly polarized microwaves, in the presence of a small magnetic field, can lead to spin-1 quantum magnetic Hamiltonians, due to the decoupling between electron spin and orbit, that is unique to the $^2Σ$ ground state molecules. We demonstrate that in one dimension, the phase diagram associated with this Hamiltonian, computed via tensor network methods, hosts the celebrated Haldane phase. We find that the Haldane phase persists even in the presence of SU(3) correction terms that break the SU(2) algebra of the Hamiltonian. We discuss the feasibility of the proposed scheme for $^2Σ$ molecules with large rotational constants such as the directly laser cooled molecule MgF for future experiments.

We study the quasinormal spectrum of massless scalar fields propagating on a family of regular BTZ black holes arising from an infinite tower of dimensionally regularized Lovelock corrections. These geometries are asymptotically AdS, reduce to the standard BTZ solution in the limit $\ell \to 0$, and resolve the central singularity by introducing a smooth core controlled by the new length scale $\ell$. The scalar quasinormal modes are computed using both Leaver's continued-fraction method and the Horowitz-Hubeny power-series method; the two approaches agree to high accuracy across the parameter space. We find that the regularization preserves linear stability ($ω_I < 0$) while qualitatively reshaping the spectrum: as $\ell$ increases, BTZ-like complex branches collide with the imaginary axis and undergo a hierarchy of bifurcations into multiple purely imaginary branches, leading to mode switching and a nontrivial reordering of overtones as functions of $\ell$ and the harmonic index $m$. Our results place regular BTZ black holes within the emerging family of bifurcating quasinormal spectra known from nearly extremal and asymptotically AdS black holes, and highlight these $(2+1)$-dimensional geometries as a controlled arena for exploring geometric mechanisms behind spectral branching and late-time ringdown in regular black hole spacetimes.

The current definition of the SI second is based on the 133Cs ground-state hyperfine transition in the microwave domain, with the most accurate realizations achieving fractional frequency uncertainties of about (1-2)E16. In contrast, state-of-the-art optical clocks now demonstrate estimated uncertainties two to three orders of magnitude lower, prompting discussion on the redefinition of the SI second. Several options for the new definition have been proposed, one of which introduces a constant N defined as the weighted geometric mean of multiple clock transition frequencies. In this work, we investigate how N can be practically realized when not all defining transitions are available and when multiple optical clocks operate with different performance levels and non-overlapping uptimes. We consider two complementary realization and reconstruction routes. One route is based on geometric-mean combinations, and the other is based on arithmetic-mean combinations. We derive consistent uncertainty expressions that incorporate both measurement uncertainties and, where required, uncertainties of recommended frequencies or frequency ratios. Using analytic three-transition case studies, we identify the parameter regimes in which each route yields a lower total uncertainty and provide explicit conditions for the crossover between them. We further address the dominant role of dead time when a hydrogen maser serves as a flywheel reference by introducing a time-segmented, time-weighted combination based on coefficient and covariance matrices, which accounts for overlapping operation and correlations across measurement intervals. Our findings offer practical guidance for minimizing total uncertainty in multi-clock realizations and contribute to ongoing efforts toward redefining the SI second.

A complete architecture for cavity-free quantum networking based on collective enhancement in Rydberg atom ensembles is presented. The protocol exploits Rydberg blockade and phase-matched directional emission to eliminate optical cavities without sacrificing performance. The architecture comprises three steps: (i) local control-ensemble entanglement via Rydberg blockade with fidelity $F_{\mathrm{gate}} \approx 99.93\%$; (ii) atom-photon conversion via Raman transitions, achieving directional emission ($η_{\mathrm{dir}} \approx 35\%$) and single-node efficiency $η_{\mathrm{node}} \approx 19\%$; and (iii) remote atom-atom entanglement via Hong-Ou-Mandel interference, producing Bell states with fidelity $F > 97.5\%$. With quantum memories enabling retry protocols, entanglement generation rates exceed $600$ Hz at 20 km separation. This cavity-free approach provides a practical and scalable pathway for distributed quantum computing and secure quantum communication.

The origin of cosmic angular momentum is a fundamental question in structure formation. We propose a novel mechanism that generates spatial angular momentum directly from quantum fluctuations during inflation. A spectator complex scalar field with global U(1) symmetry stores internal angular momentum via field-space rotation. Inflationary perturbations create spatial gradients that, upon horizon re-entry, couple to the background charge density and source a bulk momentum flow. During nonspherical gravitational collapse, this flow converts into net angular momentum. For primordial black holes forming from such collapse, the dimensionless spin can reach \(χ\sim 0.1-1\) when the small-scale power spectrum is enhanced to produce detectable abundances-far exceeding tidal torque theory predictions. This establishes a testable link between inflation, primordial perturbations, and black hole spin distributions accessible to gravitational-wave observations.

We benchmark constrained-path Monte Carlo (CPMC) on the triangular-lattice Hubbard model for several fillings and $U$ values and show that symmetry-adapted trial wave functions are essential for quantitative accuracy. Away from half-filling, simple free-electron-based trials that preserve the ground state symmetry yield energy deviations $\lesssim 1\%$ from exact diagonalization and density matrix renormalization group results. At half-filling, strong frustration in the intermediate to large $U$ regimes necessitates symmetry-projected trials to reach comparable accuracy, where both free-electron and symmetry-broken Hartree-Fock trials incur substantial constraint bias. Since the computational cost of CPMC with symmetry projection scales polynomially with system size, our results motivate its use as a practical route for studying competing ground states in strongly correlated, frustrated systems.

Projection games constitute an important class of nonlocal games where, for any answer from the first player, there is a unique correct answer for the second player. This class of games captures nonlocal games arising from constraint satisfaction problems, oracularisations, and unique games. However, due to the asymmetry between the players, projection games are in general not synchronous, and therefore the powerful results constraining the structure of almost perfect strategies for synchronous games do not apply. In this work, we adapt results of Marrakchi and de la Salle for synchronous games to show that, in both the quantum and commuting-operator models, any strategy that wins with probability $1-\varepsilon$ in a projection game gives rise to a tracial strategy that wins with probability $1-O((L\varepsilon)^{1/4})$, where $L$ is the inverse of the minimal conditional probability of a question for the second player being sampled given a question to the first. For constraint system games, this strengthens the rounding result of Paddock by eliminating the dependence on number of constraints and improving the dependence on constraint size, while also generalising to the commuting-operator setting.

Cardinality-constrained binary optimization is a fundamental computational primitive with broad applications in machine learning, finance, and scientific computing. In this work, we introduce a Grover-based quantum algorithm that exploits the structure of the fixed-cardinality feasible subspace under a natural promise on solution existence. For quadratic objectives, our approach achieves ${O}\left(\sqrt{\frac{\binom{n}{k}}{M}}\right)$ Grover rotations for any fixed cardinality $k$ and degeneracy of the optima $M$, yielding an exponential reduction in the number of Grover iterations compared with unstructured search over $\{0,1\}^n$. Building on this result, we develop a hybrid classical--quantum framework based on the alternating direction method of multipliers (ADMM) algorithm. The proposed framework is guaranteed to output an $ε$-approximate solution with a consistency tolerance $ε+ δ$ using at most $ {O}\left(\sqrt{\binom{n}{k}}\frac{n^{6}k^{3/2} }{ \sqrt{M}ε^2 δ}\right)$ queries to a quadratic oracle, together with ${O}\left(\frac{n^{6}k^{3/2}}{ε^2δ}\right)$ classical overhead. Overall, our method suggests a practical use of quantum resources and demonstrates an exponential improvements over existing Grover-based approaches in certain parameter regimes, thereby paving the way toward quantum advantage in constrained binary optimization.

We propose and analyze an enhanced hyperbolic square-root (HSQRT) deformation of the Starobinsky model in the context of $f(R)$ gravity. The original HSQRT construction provided a globally regular modification of $R^2$ inflation, curing the strong-coupling singularity at negative curvatures while preserving the characteristic exponential slow-roll plateau at large positive curvatures. Motivated by recent precision cosmological observations (such as ACT DR6 and DESI) indicating an upward shift in the scalar spectral index and a preference for deviations from pure exponential plateau behavior, we introduce a structurally minimal, quantum-motivated logarithmic enhancement. This phenomenological enhancement modifies the deep ultraviolet asymptotic regime while maintaining global tachyon-free stability, ghost freedom, and the recovery of general relativity at low curvatures to preserve standard reheating. By developing an exact parametric formulation of the Einstein frame, we demonstrate that the scalar potential of the enhanced model transitions from an exponential to an inverse-power asymptotic form, $V(φ) \simeq V_0 \left[ 1 - 6β/ (κ^2φ^2) \right]$, where the strength of this deviation is governed by a single dimensionless parameter $β$. We derive exact analytic expansions for the slow-roll observables, yielding a parameter-free leading-order spectral index $n_s \simeq 1 - 3/(2N)$ and a tunable tensor-to-scalar ratio $r \simeq 2(3β)^{1/2}/N^{3/2}$. For standard inflationary e-fold durations ($N \in [50, 60]$), this model drives the spectral index directly into the newly favored observational window ($n_s \simeq 0.970\text{--}0.975$) and predicts an exceptionally small running $α_s \simeq -3/(2N^2) \in [-0.00060, -0.00042]$, while providing a viable, parameter-controlled target for next-generation cosmic microwave background observatories.

We present a toolkit, CosmoDS, designed to study cosmological models at the background level using dynamical system analysis within the Cobaya framework. Dynamical system analysis is a powerful mathematical approach for studying nonlinear systems and is widely used in cosmology to investigate the stability and evolution of different cosmological models, particularly those involving dark energy. In this code, we provide a framework for constraining cosmological models using a dynamical system formulation. Most importantly, the toolkit is directly integrated with the Cobaya interface, allowing users to take advantage of the sophisticated statistical and inference tools already implemented in Cobaya for cosmological parameter estimation and model analysis.

We investigate the properties of neutron stars when a massive scalar field, which could comprise all dark matter, is non-minimally coupled to the Ricci scalar. This coupling generates additional contributions to the field's effective mass, leading to tachyonic instabilities inside neutron stars and giving rise to rich phenomenology. Within this framework, we obtain neutron-star configurations with maximum masses exceeding 2 $M_\odot$, even when hyperons, which typically soften the equation of state and significantly lower the maximum mass, are included. Furthermore, we find that larger coupling strengths lead to multiple solutions for the scalar-field configuration. We analyze the structure of the corresponding effective potential responsible for this behavior. We also investigate how the inclusion of a scalar self-interaction term, in addition to the non-minimal coupling, modifies the resulting neutron-star properties.

We establish a general no--go theorem demonstrating that all traversable wormhole configurations in Unimodular Gravity necessarily require exotic matter. The proof relies solely on the geometric flaring-out condition, $b'(r_0) \leq 1$, which directly implies that $ρ(r_0) + p_r(r_0) \leq 0$ at the throat. This condition represents a violation of the Null Energy Condition and, consequently, of the Weak and Strong Energy Conditions, independently of the particular choice of shape function, redshift function, or equation of state. This result holds for both tidal and zero-tidal-force configurations, showing that the requirement of exotic matter is a fundamental geometric consequence of the traversability condition rather than an artifact of specific solution choices. Therefore, Unimodular Gravity shares this fundamental constraint with General Relativity.

Real-time dynamics of quantum observables provide direct access to excitation spectra and correlation functions in quantum many-body systems, but currently available quantum devices are limited to short evolution times due to decoherence. We propose a neural ordinary differential equation (Neural ODE) framework with physics-driven designs to reconstruct long-time operator dynamics from short-time measurements. By expanding observables in the Pauli basis and exploiting locality and symmetry constraints, the operator evolution is reduced to a tractable set of coefficients whose dynamics are learned from data. Applied to the transverse-field Ising model, the method accurately extrapolates long-time behavior and resolves excitation spectra from noisy short-time signals. Our results demonstrate a scalable and data-efficient strategy for extracting dynamical and spectral information from practical quantum hardware.

In realistic hardware for quantum computation that possesses fault-tolerance, non-stationary noise and stochastic drift lead to logical failure from the temporal accumulation of errors, not from independent events. Static decoding and fixed calibration techniques are structurally incompatible with this situation because they do not take into account temporal correlations between errors or control-induced back-action of errors. These effects motivate control policies that must track noise evolution across correction cycles, rather than respond to individual syndromes in isolation. We treat fault-tolerant quantum computation as a stochastic control problem, modelled using reduced quantum dynamics in which Pauli error processes are governed by latent noise parameters that vary temporally. From this perspective, logical failure arises through the accumulation of a hazard variable, and the corresponding control objective depends on the full history of observations. Operating under these conditions, a Chronological Deep Q-Network (Ch-DQN) maintains an internal belief state that tracks both noise evolution and accumulated hazard. During training, backward refinement of trajectories is used to sample slowly drifting modes of operation, while runtime inference remains strictly causal. A fractional meta-update stabilizes learning in the presence of non-stationary, control-coupled dynamics. Through multi-distance simulations that incorporate stochastic drift and feedback from decision-making, Ch-DQN suppresses hazard accumulation and extends logical survival time relative to static and recurrent baselines. Error correction in this regime is therefore no longer a static decoding task, but a control process whose success is determined over time by the underlying noise dynamics.

We introduce a sparse classical representation, a truncation strategy and a shot-efficient sampling method to push the classical prediction of quantum error correction thresholds beyond Clifford operations and Pauli errors. As two illustrations of the potential of our method, we first show that coherent noise error thresholds, when computed at the circuit level (i.e taking into account full syndrome circuits) for distances up to d=9, are systematically overestimated (by a factor of about 4) by a Pauli-twirling approximation of the noise. We then apply our method to the recently introduced magic-state cultivation protocol. We show, through shot-efficient importance sampling, that, at distance d=5, the multiplicative factor between the T-gate and the S-gate injection error rate is not the one conjectured from low-d computations: it can be as large as 7.

Lieb and Ruskai's strong subadditivity theorem, which shows that the conditional mutual information must be nonnegative, is fundamental in quantum theory. It has numerous applications, such as in quantum error correction. When the mutual information is zero, the Petz recovery map can be used to reconstruct the quantum channel. When the mutual information is small, one seeks to define an optimal recovery channel. To this end, a mathematical characterization of the mutual information is desirable. We address this problem by providing an exact characterization of the mutual information, along with characterizations for other entropies. Our controls are sharp, leaving no room for improvement, in the sense that we provide equalities, regardless of whether the mutual information (or remainder) is small or large. We transform the definitions of these entropies into a summation of explicitly constructed terms, and the definition of each term obviously demonstrates the desired positivity/convexity/concavity. The summation converges rapidly and absolutely in a chosen elementary norm.

We introduce new parallel algorithms for efficiently simulating stabilizer (Clifford) circuits on GPUs, with a focus on data-parallel tableau evolution and scalable handling of projective measurements. Our approach reformulates key bottlenecks in stabilizer simulation -- such as Gaussian elimination and measurement updates -- into GPU-tailored primitives that eliminate sequential dependencies and maximize memory coalescing. We implement these techniques in QuaSARQ, a GPU-accelerated stabilizer simulator designed for large qubit counts and many-shot sampling. Across a broad benchmark suite reaching 180,000 qubits and depth 1,000 (roughly 130M gates), QuaSARQ shows substantial runtime improvements, with up to 105$\times$ speedup, and over 80% energy reduction on demanding instances. Moreover, QuaSARQ consistently outperforms Stim, a state-of-the-art CPU-optimized stabilizer simulator, as well as Qiskit-Aer (CPU/GPU), Qibo, Cirq, and PennyLane. Finally, QuaSARQ exhibits a significant advantage in many-shot sampling on large workloads. These results demonstrate that our parallel algorithms can significantly advance the scalability of stabilizer-circuit simulation, particularly for workloads involving extensive measurements and sampling.

We evaluate calibration-based digital twins for IBM Quantum hardware, aiming to reproduce hardware measurement outcomes on classical simulators. We present a workflow that builds twins from downloadable calibration CSV files by mapping coherence times, gate and readout error rates, and operation durations to thermal-relaxation, depolarizing, and readout error channels, while reconstructing a directed coupling map to restore connectivity constraints during transpilation. We compare four twin variants (CSV-built, backend-derived simulator, backend-derived noise model, and fake-backend snapshots) under a common execution and validation protocol. Experiments on two IBM QPUs, ibm_brisbane and ibm_sherbrooke, use randomized five-qubit circuits of depths 10, 20, and 30 across four optimization levels. Weighted Jaccard similarity indicates that twins constructed from downloadable calibration CSV data often achieved the closest agreement with hardware, while backend-derived twins provided competitive and practical baselines. The results further show that agreement depends on both the target device and the transpilation settings, underscoring the need to validate digital twins for the specific execution setup rather than assuming transferability across devices.

In three spacetime dimensions, pure Einstein gravity admits no local propagating degrees of freedom, yet nontrivial gravitational backgrounds such as the BTZ black hole provide a natural arena to probe dynamical extensions of the theory. In quadratic $f(R)$ gravity the Ricci scalar becomes a propagating degree of freedom - the scalaron. We investigate how localized Maxwell-Higgs vortices excite this scalar mode in a static BTZ black-hole background. Working in the perturbative regime $α\ll \ell^2$, the trace equation reduces to a massive Klein-Gordon equation for the curvature scalar sourced by the trace of the vortex energy-momentum tensor. Using the Sturm-Liouville structure of the radial operator, we construct the corresponding Green function and obtain the curvature profile generated by an arbitrary localized source. The induced excitation exhibits a universal asymptotic decay $R(r) \sim r^{-(1+ν)}$, independent of the detailed vortex structure. The scalar excitation is linearly stable, carries finite energy, and produces parametrically suppressed backreaction, ensuring the smooth recovery of the Einstein limit. These results provide a concrete realization of how higher-curvature corrections activate the unique local gravitational degree of freedom in three dimensions and how localized sources excite this scalar mode in black-hole spacetimes.

Minimizing the energy of a many body system tends to favor order, but classical frustration and quantum fluctuations destabilize that order. The tension between these effects can produce exotic quantum states of matter. Quantum spin liquid (QSL) states emerge in models of localized magnetic moments where the crystal lattice connectivity frustrates ordering, and the exchange interaction of neighboring spins strengthens quantum fluctuations. Experimentally identifying a QSL in a real material is challenging from the lack of an order parameter. Piecing together evidence from varied techniques is necessary for diagnosing the nature of the ground state -- QSL or otherwise -- of a frustrated spin system. In this work, we use coplanar superconducting resonators to probe magnetic excitations in epitaxially grown thin films of a spin liquid candidate TbInO3. Adapting microwave techniques from the field of circuit quantum electrodynamics, we measure responses of these thin films whose volume is too low for applying conventional bulk techniques. In-plane susceptibility extracted from the spin resonance signal indicates extreme frustration of magnetic order down to 20 mK, over two orders of magnitude lower than the Curie-Weiss energy scale. Through a crystal field analysis, we identify the doublet eigenstates comprising the ground state. As a consequence of improper ferroelectricity, Tb moments split into two flavors with distinct g-factors reflecting the local crystal field environment of each site. Spin-orbit coupling, crystal fields, magnetic frustration and improper ferroelectricity distinctively combine to shape the magnetic ground state of TbInO3. This work establishes a measurement technique using superconducting resonators to probe thin films of frustrated magnets, and applies this technique towards building a coherent understanding of the magnetic properties of TbInO3.

Reversible computation requires that intermediate data be explicitly undone rather than discarded. In quantum programming, this principle appears as uncomputation, usually treated as a technical cleanup mechanism. We instead present uncomputation as a semantic foundation. In the Qutes language, we introduce a formal model of \emph{Scope-Bounded Liveness-Guided Uncomputation}, where lexical scope bounds variable lifetime and static liveness and entanglement analysis determine the earliest safe reclamation point. We define semantic lifetime and a Restoration Invariant ensuring that temporary quantum information disappears once it becomes semantically irrelevant. We prove compositional correctness under nested scopes and show that early reclamation can reduce circuit depth by avoiding critical-path overhead and can bound peak live qubits through disciplined ancilla reuse. Finally, we show that parameter passing semantics emerges from the same lifetime discipline, with pass-by-value and pass-by-reference corresponding to different lifetime boundaries, and we characterize the constraints (irreversibility, persistent entanglement, and aliasing) under which automatic uncomputation must be restricted.

Spin chains have been widely studied as quantum channels for short-distance communication in quantum devices, where many-body dynamics can mediate quantum-state transfer between distant sites. In finite unmodulated chains, however, dispersion and interference effects associated with the static Hamiltonian often limit the achievable transfer fidelity. Here we investigate the transfer of single-qubit and Bell states in finite $XX$ spin chains under periodic switching between two Hamiltonians with different boundary couplings. Inspired by Parrondo's paradox, we examine whether alternating between two configurations that individually yield suboptimal transfer fidelities can generate enhanced coherent transmission. Using Floquet theory together with numerical simulations in the single-excitation subspace, we show that periodic driving can outperform static configurations and achieve higher transfer fidelities. This enhancement originates from the noncommutativity of the driven Hamiltonians and reflects a purely coherent interference effect. We further analyze the dependence of the protocol on system size and driving parameters and examine its robustness to asymmetric boundary couplings. Our results show that the transfer fidelity remains stable under moderate disorder, indicating that simple time-dependent control of boundary couplings provides an effective strategy to enhance quantum-state transfer in spin-chain communication channels and optimize quantum information processing in engineered many-body systems.

We present simple and practical strategies to reduce the variance of Monte Carlo estimators. Our focus is on variational Monte Carlo calculations of atomic forces and pressure in electronic systems, although we show that the underlying ideas apply more broadly to other observables, like pair-correlation and angular-distribution functions, and other methods, including molecular dynamics. For Pulay-type contributions, we show that a minor modification based on the Metropolis acceptance ratio softens the power-law divergence of the variance to a logarithmic one, and that inexpensive regularizations can further suppress outliers at the price of a controlled small bias. For Hellmann-Feynman forces, we derive compact variance-reduced estimators for periodic systems that are straightforward to implement in standard Monte Carlo codes. The approach is illustrated for high-pressure metallic hydrogen with more than a hundred atoms described by neural quantum states, including an application to molecular dynamics driven by the improved forces.

We reconsider the late-time evolution of galaxy peculiar velocities in the 1+3 covariant approach to cosmological perturbation theory. It has recently been claimed that this approach predicts substantially stronger growth of peculiar velocities than standard metric-based perturbation theory -- on the grounds that the covariant treatment is fully relativistic whereas standard treatments are effectively Newtonian. We show that this is not the case. When the covariant equations are applied consistently, the $1+3$ approach reproduces exactly the standard perturbative result for peculiar-velocity growth. The stronger growth laws claimed in recent work arise from an inconsistent treatment of the coupled covariant system, in which terms constrained by the field equations are treated as if they were independent sources. Further claims are made that the stronger bulk flows can mimic accelerated expansion in a dust universe. We argue that these claims rest on a confusion between the kinematics of an arbitrarily chosen observer congruence and the physical expansion of the matter congruence traced by galaxies. We conclude that the standard treatment of peculiar velocities is correct and fully relativistic~-- and does not lead to anomalous bulk flows or to apparent accelerated expansion.

We construct a two-dimensional action on the celestial sphere that describes the infrared sector of Abelian gauge and gravitational theories in four dimensions. In particular, we use the holographic model to reproduce (1) antipodal matching conditions for the superphaserotation and supertranslation Goldstone modes in four dimensions, (2) leading soft photon and graviton theorems, and (3) infrared factorization of amplitudes with generic dressed $in$ and $out$ states. Using (3), we reproduce the infrared divergences that plague the standard undressed amplitudes, and show that amplitudes involving Faddeev-Kulish dressed states are infrared finite. As a corollary, we use our holographic model to construct an infinite class of dressed states that give rise to infrared finite scattering amplitudes.

Nonlocal RT gravity has proven effective in explaining the late-time cosmic acceleration while remaining consistent with local gravity tests. However, most previous cosmological studies of this theory have assumed an isotropic background, which may not fully capture the slight anisotropies suggested by current observations, such as those inferred from Type Ia supernovae data. In this paper, we investigate the dynamical evolution of an anisotropic Bianchi type I universe within the framework of nonlocal RT gravity. By introducing six dimensionless variables, we construct the corresponding dynamical system and perform a detailed phase-space analysis. An unexpected finding is that, contrary to many dark energy models and modified gravity theories in which anisotropies decay with time, nonlocal RT gravity predicts a growth of cosmic anisotropy. This behavior poses a challenge to the cosmic no-hair theorem within the nonlocal RT gravity scenario.

Accurately estimating observables on noisy quantum devices remains a central challenge for near-term quantum algorithms. While quantum error mitigation techniques can reduce noise-induced bias, they often rely on unverifiable assumptions about the circuit noise, and cannot guarantee the magnitude of residual bias error. Here, rather than using classical resources to mitigate a noisy quantum circuit execution, we propose a hybrid algorithm that uses quantum resources to improve the accuracy of approximate classical Pauli-path simulation. Our protocol, Quantum Enhanced Pauli Propagation (QuEPP), uses Clifford perturbation theory (CPT) to construct a classically simulable ensemble of Clifford circuits from the low-order terms in CPT, which directly provide the approximate classical Pauli-path simulation of the target circuit. Noisy quantum expectation values of this ensemble are then used to infer a global rescaling factor that corrects quantum execution of the target circuit, providing higher-order contributions absent from the truncated low-order classical simulation. This approach requires no noise characterization, applies to arbitrary circuits, and provides a provable route to asymptotically unbiased estimates. Using IBM Heron hardware, we demonstrate QuEPP on 2D random mirror circuits of up to 49 qubits and circuit depth 80, as well as Trotterized Hamiltonian evolution, showing consistent improvements beyond classical CPT and unmitigated quantum results. QuEPP offers a simple, scalable, and model-free framework for enabling accurate quantum computation in the pre-fault-tolerant era.

I introduce a novel realistic, stochastic approach to quantum mechanics by extending the recently proposed grabit formalism \cite{braun_stochastic_2022} to two Twin Worlds. According to the picture developed, we live at the intersection of two worlds with identical stochastic laws of evolution. Our World is limited to that intersection, and only coincidence events from the two Twin Worlds, post-selected automatically by our restriction to the intersection, have physical reality in Our World. This fully reproduces standard non-relativistic quantum mechanics, including Born's rule and the violation of Bell's inequality. I derive the stochastic evolution equation in each Twin World that fully reproduces Schrödinger's equation for an arbitrary number of particles with arbitrary interactions, and demonstrate that hall-mark quantum effects such as tunnling are correctly reproduced.

The relentless scaling of classical microelectronics has been enabled by the precision and reproducibility of deep-ultraviolet (DUV) optical lithography. Implementing large-scale superconducting quantum processors will require cryogenic microwave components that follow a similarly scalable fabrication path. This need is particularly acute for high circuit-density devices such as traveling-wave parametric amplifiers (TWPAs), where recent implementations have demonstrated high gain, broad bandwidth, high saturation power, and near-quantum-limited noise, but trade-offs between footprint, insertion loss, and scalable integration remain. Here, we demonstrate a four-wave-mixing TWPA fabricated via a hybrid scheme that combines DUV-defined planar circuit elements with electron-beam-patterned Josephson junctions, constituting a first step toward fully scalable manufacturing. The device combines a compact footprint with broadband gain from 3 to 11 GHz and an average 1 dB compression point of -102 dBm. By using planar capacitors to reduce loss, it operates near the quantum limit, with added noise near 0 and 1.5 photons above the standard quantum limit and an average of 0.4 photons in the 4 to 8 GHz band. The phase-matching stopband remains narrow, with a bandwidth of 43 MHz, consistent with resonator-frequency variation below 1% and indicative of the uniformity enabled by DUV lithography. These results show that DUV-defined planar elements can enable compact, low-loss, near-quantum-limited TWPAs and provide a promising route toward high-density cryogenic microwave hardware for large-scale quantum systems.

We introduce Neural Tensor Network States ($ν$TNS), a variational many-body wave-function ansatz that integrates deep neural networks with tensor-network architectures. In the $ν$TNS framework, a neural network serves as a disentangler of the wave-function, transforming the physical degrees of freedom into renormalized variables with much less entanglement. The renormalized state is then efficiently encoded by a back-flow tensor network. This construction yields a compact yet highly expressive representation of strongly correlated quantum states. Using convolutional neural networks combined with matrix product states as a concrete implementation, we obtain state-of-the-art variational energies for the spin-$1/2$ $J_1$-$J_2$ Heisenberg model on the square lattice at the highly frustrated point $J_2/J_1=0.5$, for systems up to $20\times 20$ with periodic boundary conditions. Finite-size scaling of spin, dimer, and plaquette correlations exhibits power-law decay without magnetic or valence-bond long-range order, consistent with a gapless quantum spin-liquid ground state at that point.This $ν$TNS framework is flexible and naturally extensible to other neural and tensor-network structures, offering a general platform for investigating strongly correlated quantum many-body systems.

In this paper, we investigate the dynamics of massive particles and the associated gravitational waveforms in the spacetime of a black hole within the framework of Einstein-Bumblebee gravity. Our analysis encompasses both charged and uncharged black hole configurations, with a particular focus on the spontaneous Lorentz symmetry breaking mechanism inherent to this model, which is governed by a dimensionless coupling parameter $l$. We analyze the geodesic equations and the effective potential to determine the allowed parameter space for bound orbits, demonstrating that in the charged case, both the Lorentz-violating parameter $l$ and the electric charge $Q$ significantly enhance the confinement capacity of the potential, thereby broadening the energy and angular momentum windows for bound states. A key focus is placed on the classification and properties of periodic orbits, characterized by rational frequency ratios using the whirl, zoom, and vertex taxonomy. We demonstrate that in the uncharged case ($Q=0$), the radial effective potential and standard innermost stable circular orbit (ISCO) properties are degenerate with those of a Schwarzschild black hole. However, despite this degeneracy in static potential properties, the structure of periodic orbits exhibits qualitative differences, providing a possible observational signature that can break this degeneracy. Finally, we compute the corresponding gravitational waveforms extracted from these periodic orbits using the quadrupole formula. The results reveal that $l$ and $Q$ introduce contrasting phase-shifting effects on the waveforms. This suggests that bumblebee gravity leaves measurable imprints on gravitational-wave signals that could be detected by future space-based gravitational-wave observatories.

We will prove that there is a direct relationship between the Poincare subgroup of translations, and the group of tetrad transformations LB1 introduced in a previous manuscript. LB1 is the group composed by SO(1; 1) plus two kinds of discrete transformations. Translations have been extensively studied under the scope of gauge theories. By using the geometric structures built to prove this elementary result we will generalize it to the case of what we might call local translations. A special case of the latter is the Bondi-Metzner-Sachs subgroup of supertranslations. In order to accomplish this goal and since the group of translations is four-dimensional we will prove first that it is isomorphic to (x LB1)4. In order to prove this claim we will introduce a system of differential equations involving several kinds of fields. Abelian, non-Abelian, spinor, gravitational. These fields will constitute the structure needed to build local tetrads of a new kind that allow for the proof to be carried out with simplicity. Results already obtained involving similar but not equal tetrads will be useful in our constructions and demonstrations. Translations and generalized translations isomorphic to tensor products of LB1 groups are not trivial results. Because the LB1 group is composed by SO(1; 1) and two discrete transformations.

Quantum synchronization (QS) in dissipative systems is often inferred from smooth phase locking, leaving open whether its breakdown constitutes a genuine nonequilibrium transition. Here we introduce a Liouvillian framework that classifies driven-dissipative dynamics by the structure of the undriven dissipative background and show that QS breaks down via a Hopf-type dynamical phase transition into a boundary time crystal (BTC). The character of this transition is determined by the background attractor: systems with a self-sustained oscillator (SSO) support robust non-resonant BTCs, whereas those with a polar fixed point (PFP) sustain BTCs only at resonance and lose them under detuning. We identify sharp dynamical and spectral signatures of the QS-BTC transition and thereby establish, within U(1)-symmetric collective-spin Lindbladians driven by a single coherent tone, a background-based allowed/forbidden criterion that unifies QS, its breakdown, and time-crystalline order within a single Liouvillian framework.

We propose a Floquet-engineered framework for the coherent control of the light-matter interaction in a two-level system (TLS) located in a time-modulated cavity. Strictly phase-preserving operation of the TLS-cavity interaction is demonstrated, allowing the interrupt and retrieval of coherent Rabi oscillations without the loss of quantum information. By introducing a phonon reservoir, it is proved that the frequency instability induced from non-Markovian processes does not produce significant phase decoherence during Floquet modulation. Our results provide new insights into the fundamental physics of a driven quantum system and establish Floquet engineering as a powerful tool for coherent quantum information processing.

Given random walk on a graph, the corresponding discrete-time quantum walk can be constructed using the method proposed by Szegedy. On the other hand, given a partition of the set of states of a Markov chain, one can study the corresponding aggregated process. We extend the aggregation technique to the level of quantum Markov chains. We provide conditions under which application of these two operations - Szegedy's quantization and aggregation - give the same result. In particular, we show that the conditions are satisfied in the case of the random walk on graphs equipped with equitable partitions. We present several examples, which include the classical/quantum walks on Platonic solids. We discuss also relation of discrete-time classical/quantum walks on $N$-dimensional hypercube and the Ehrenfests urn model with $N$ particles. We apply our technique for of discrete-time walks on Cayley graphs of free groups. We also compare our results with those obtained using Cantero-Moral-Velazquez uniformization of unitary matrices.

Controlled gates are key components in various quantum algorithms. Improving on the prior work of Gosset et al., we show that, for an allowed error $\varepsilon$, $3\log_2(1/\varepsilon) + o(\log(1/\varepsilon))$ $T$ gates are sufficient to approximate most multi-qubit controlled SU(2)s. We also show that this T-count matches the lower bound when the use of an almost controlled gate is prohibited. As an application, general controlled gate synthesis and efficient SU(4) gate synthesis are given.

Frequency-resolved photon emission is central to applications from quantum information encoding to high-resolution spectroscopy, and then studying their correlations is therefore essential for revealing the underlying emission pathways and multiphoton statistics. Here, we investigate frequency-resolved N-photon correlations in an ultrastrongly coupled cavity QED system where a qubit interacts with a single-mode cavity. Owing to counter-rotating interactions, the eigenstates and energy spectrum are strongly modified, giving rise to rich spectral and statistical properties in the emitted frequency-resolved photons. Through frequency-selective detection, we reveal pronounced multiphoton antibunching, as well as multiphoton bunching originating from cascade transitions among dressed eigenstates. In particular, we show that parity symmetry plays a decisive role in shaping these correlations. The symmetry-breaking opens additional transition channels and dramatically enhances the generation of correlated photon pairs and even photon triplets of different frequencies. Our work extends frequency-resolved correlations to the ultra-strong coupling regime and demonstrates their potential as a sensitive probe of symmetry in light-matter interaction systems.

We present CETOmega, a unified framework that completes gravity through a causal-informational principle. The theory reconciles general relativity and quantum mechanics within a strictly four-dimensional, nonlocal, and causal formulation. At its core lies an analytic and retarded kernel K^-1(Box_R), derived from a discrete causal network, which governs the propagation of the gravitational and scalar sectors. A scalar field, the texon, emerges as the effective excitation of causal connectivity and accounts simultaneously for dark matter and dark energy without introducing extra degrees of freedom or breaking locality. The formalism ensures analyticity, spectral positivity, and holographic completeness. The kernel admits a Stieltjes representation with positive spectral density rho(mu) greater than or equal to zero, guaranteeing unitarity and causal propagation. At cosmological scales, CETOmega predicts stable inflationary dynamics consistent with Planck observations. Black hole ringdown frequencies acquire perturbative corrections controlled by the causal scale and remain subleading for astrophysical black holes within the fiducial window l* between 10^-5 and 10^-4 meters, where l* defines the mean causal correlation length of the texonic field. CETOmega thus provides a complete, causal, and informational foundation for spacetime dynamics, recovering Einstein gravity in the infrared while extending its validity to the quantum and cosmological domains.

We present results from an all-sky search for continuous gravitational waves, using three different methods applied to the first eight months of LIGO data from the fourth LIGO-Virgo-KAGRA Collaboration s observing run. We aim at signals potentially emitted by rotating, non-axisymmetric isolated neutron star in the Milky Way. The analysis spans a frequency range from 20 Hz to 2000 Hz and accommodates frequency derivative magnitudes up to $10^{-8}$ Hz/s. No statistically significant periodic gravitational wave signals were detected. We establish 95% confidence-level (CL) frequentist upper limits on the dimensionless strain amplitudes. The most stringent population-averaged strain upper limits reach 9.7 $\times$ $10^{-26}$ near 290 Hz, matching the best previous constraints from 250 to $\sim$1700 Hz while extending coverage to a much broader spin-down range. At higher frequencies, the new limits improve upon previous results by factors of approximately $\sim$1.6. These constraints are applied to three astrophysical scenarios: 1) the distribution of galactic neutron stars as a function of spin frequency and ellipticity; 2) the contribution of millisecond pulsars to the GeV excess near the galactic center; and 3) the possible dark matter fraction composed of nearby inspiraling primordial binary black holes with asteroid-scale masses.

We obtain exact charged AdS black hole solutions in Einstein Lambda gravity including the effects of Born Infeld nonlinear electrodynamics and Perfect Fluid Dark Matter. The influence of the PFDM and BI parameters on the event horizon is analyzed. We compute the conserved and thermodynamic quantities and verify that they satisfy the first law of thermodynamics. Thermal stability is studied in the canonical ensemble using the heat capacity and Helmholtz free energy showing how PFDM and BI parameters affect local and global stability regions. We further investigate the thermodynamics in the extended phase space by treating the cosmological constant as thermodynamic pressure obtaining consistent conserved quantities and confirming the first law. The Ehrenfest equations are analytically verified demonstrating that the critical behavior corresponds to a second order phase transition. Heat engines associated with these black holes are also constructed to examine how PFDM and BI parameters influence their efficiency. Finally we analyze the geodesic structure through timelike and null trajectories using the effective potential determining conditions for stable and unstable circular orbits the innermost stable circular orbit and the photon sphere. PFDM significantly modifies the orbital structure while BI corrections are weaker.

Over the past decade, the Python-based Simulations of Chemistry Framework (PySCF) has developed into a widely used open-source platform for electronic structure theory and quantum chemical method development. This article reviews the major advances since the previous overview in 2020, covering new modules and methodology, infrastructure changes, and performance benchmarks.

Characterization of the local spin environment of single diamond nitrogen-vacancy centers is a critical task for quantum sensing, quantum networking, and diamond materials optimization. We introduce NVRNet, a physics-informed simulation-to-reality pipeline that maps a fast acquisition, noisy Ramsey photoluminescence (PL) trace to a denoised waveform as well as outputting a direct estimate of hyperfine coupling to ${}^{13}\mathrm{C}$ spins in the environment. The denoiser is a two-stage time-frequency UNet followed by an attention-augmented time-domain UNet, pretrained on Hamiltonian-based simulations with experimentally calibrated noise. The simulation-pretrained, experimentally fine-tuned denoiser reduces the median reconstruction error on held-out few-sweep experimental traces to $0.44$-$0.67\times$ that of the raw experimental noisy traces across the three NV centers. A transformer-based estimator trained on simulation labels then predicts hyperfine parameters, and forward reconstruction from the inferred parameters reproduces the dominant experimental time- and frequency-domain features, with representative normalized FFT reconstruction errors of 0.10-0.19. These results establish NVRNet as a fast, hardware-compatible route to hyperfine inference from minimal Ramsey data.

In this paper, we present a statistical model of spacetime trajectories based on a finite collection of paths organized into a branched manifold. For each configuration of the branched manifold, we define a Shannon entropy. Given the variational nature of both the action in physics and the entropy in statistical mechanics, we explore the hypothesis that the classical action is proportional to this entropy. Under this assumption, we derive a Wick-rotated version of the path integral that remains finite and exhibits both quantum interference at the microscopic level and classical determinism at the macroscopic scale. In effect, this version of the path integral differs from the standard one because it assigns weights of non-uniform magnitude to different paths. The model suggests that wave function collapse can be interpreted as a consequence of entropy maximization. Although still idealized, this framework provides a possible route toward unifying quantum and classical descriptions within a common finite-entropy structure.

We characterize single-mode vacuum squeezing generated by a SNAIL Parametric Amplifier (SPA) operated under conditions representative of practical sensing and qubit-readout experiments. Motivated by prior expectations that Kerr-induced distortion limits squeezing in degenerate parametric amplifiers, we varied external flux and pump power to explore operating points where Kerr nonlinearity is theoretically minimized. We find that for practical applications where the squeezing frequency is fixed, the Kerr was variable by about a factor of two and the achievable squeezing showed no significant dependence on Kerr. Theoretical modeling supports this observation and indicates that baseline Kerr values in state-of-the-art SPAs are already too small to impose a practical limitation. Instead, squeezing was dominated by internal resonator loss and insertion loss in the microwave chain. These results indicate that, in practical SPAs, reducing loss, rather than suppressing Kerr, is the primary route to improved squeezing performance.

Hydrogen contamination in Josephson junctions is a potential source of device-to-device variability and two-level-system loss in superconducting qubits. In this work, we investigate hydrogen incorporation in oxidized aluminum barriers by combining molecular dynamics simulations with atomistic quantum transport calculations. The oxide growth simulations are performed using CHGNet for Al surfaces exposed to dense O$_{\text{2}}$ and H$_{\text{2}% }$O environments, yielding amorphous AlO$_{\text{x}}$ layers with hydrogen content comparable to experimentally relevant levels. From $400$ statistically independent samples, we find that the number of H atoms in the oxide is well described by a beta-binomial distribution, reflecting correlations induced by the self-limiting oxidation process. Structural analysis shows that most hydrogen atoms reside near the AlO$_{\text{x}}$ surface and predominantly form Al-OH and Al-OH-Al motifs. To assess the impact of hydrogen on transport, we construct Al/Al$_{\text{2}}$O$_{\text{3}} $/Al junction models and perform NEGF-DFT calculations with NanoDCAL, using a GGA+U scheme to calibrate the band gap and band alignment. H atoms are found to increase the transmission coefficient near the Fermi level and shift the electronic structure in a manner consistent with effective p-type doping. By combining the H atom number statistics from molecular dynamics with the transmission coefficients from quantum transport calculations, we obtain a probability distribution for the Josephson energy. For a Josephson junction with an average hydrogen content of $2.56$ at.\%, the resulting Josephson energy is predicted to be $% E_{J}/h=10.92\pm 0.26$ GHz. These results provide an atomistic picture of hydrogen contamination and an estimate of device variability in Josephson junctions.

Phase estimation with potentially large phase values, i.e., with large dynamic range, has many applications in quantum metrology, for example to atomic clocks. A recently proposed phase estimation scheme approaches the Heisenberg scaling in this global setting using sequences of increasingly squeezed Gaussian states as probes and adaptively chosen, potentially mid-circuit, measurements. In this work, we first observe that the pattern of increase in the squeezing of the probes is applicable even to states with some non-Gaussian features. We then propose an experimentally feasible version of this phase estimation scheme, based on the alternating application of one-axis twist (OAT) operations and rotations. Our protocols are explicitly described in terms of multiple OAT angles whose durations decrease polynomially with system size and spin-squeezing parameters that decay as $N^{-μ}$, with $μ>2/3$ in most cases. Using numerical computation of the system-size dependence $N^{-ν}$ of the Bayesian mean-squared error of an estimator, we show that these states are suitable for use in the phase estimation scheme, and highlight the protocols to achieve $ν=17/9$ and $53/27$ using two and three OAT operations respectively in the last adaptation stage. We also analyze the limited non-Gaussianity of the resulting probe states and discuss the role of non-Gaussianity in this protocol more generally. Finally, we analyze how robust these protocols are with respect to imperfections such as particle number fluctuations and coherent control fluctuations.

We compute the one-loop QED $β$-function coefficient directly from heat kernel data of the twisted Spin$^c$ Dirac operator on $S^3 \times S^1$. Using $ζ$-function regularization, the logarithmic scale dependence is encoded in the $a_4$ coefficient of the spectral expansion. The $F_{μν} F^{μν}$ term in $a_4$ yields exactly $β(e) = e^3/(12π^2)$, independent of $r$, $L$, or background, verifying spectral RG flow without flat-space propagators. The result is independent of the radii of $S^3$ and $S^1$ and of the choice of gauge background, providing a parameter-free consistency check that spectral data on compact manifolds encode renormalization group information. Beyond a mere verification of the coupling flow, this result serves as a non-trivial consistency check of the Spectral Action Principle in a curved background. It demonstrates that universal quantum corrections can be extracted purely from geometric spectral invariants, distinguishing this geometric spectral derivation from momentum-space propagator methods.

The theory of Force-Free Electrodynamics (FFE) provides a robust framework for modeling the magnetospheres of compact objects, where the electromagnetic field's energy density dominates the surrounding plasma. Central to this theory is the existence of two-dimensional integral submanifolds, or field sheets, which foliate the spacetime. While it is established that every null force-free field possesses an associated 2-D null geodesic foliation, the converse, identifying which null geodesic congruences can support a force-free solution, remains a non-trivial computational challenge. In this paper, we extend the foliation-based approach to null FFE by addressing two primary obstacles to the existence of a solution: the equipartition of null mean curvature and the involutivity of the field sheet distribution. We prove a general existence theorem demonstrating that for any given null geodesic congruence, there always exists a local rotation of a 2-D basis transverse to the geodesic congruence that satisfies the equipartition condition. Furthermore, we establish that a shear-free null geodesic congruence is sufficient to guaranty the existence of an arbitrary function of three variables such that any choice of such a function will generate a null field sheet foliation. Additionally, each unique foliation will be associated with a null force-free field that further contains an arbitrary function of two variables. These results are formally linked to the vanishing of the shear tensor, providing a coordinate-independent geometric criterion for the existence of null FFE solutions. We illustrate these theorems with explicit examples in Schwarzschild and Kerr geometries and present new, non-trivial exact null solutions in flat spacetime and for the C-metric.

The Alicki-Lindblad-Fannes dynamical (ALF) entropy measures the rate at which new information is gathered about a quantum system by inspecting its long-time evolution. We propose an extension of the ALF entropy to open quantum dynamics as a measure of back-flow of information from the environment. Such a proposal is stronger than the existing ones based only on the open system reduced dynamics. In the case of a qubit collisionally coupled to a classical spin chain, we obtain an exact expression for the $\textit{open-system ALF entropy}$ explicitly depending on the environment correlations. An extreme case shows how the information flow from environment to system corresponds to vanishing entropy production as for reversible finite quantum systems.

Accessing the physical mechanisms behind non-Markovian phenomena in open quantum dynamics requires the study of the statistical properties of the joint system-environment dynamics. This is impossible at the level of the reduced dynamics of the open system alone as the latter is obtained by suitably eliminating the environment. The task is instead made possible by considering multi-time correlation functions involving observables of the open system, only: the open system-environment interactions turn them into global ones thus building up correlations between the two systems. Multi-time correlations form the basis of both the theory of quantum stochastic processes and of the Alicki-Lindblad-Fannes dynamical entropy (ALF entropy for short). This latter quantity provides for quantum systems a measure of the dynamical entropy production as the Kolmogorov-Sinai entropy does for classical systems. In the case of a collisional model whereby the dissipative dynamics of a finite-level system is obtained by its coupling to an infinite classical spin chain, the ALF entropy can be explicitly computed. It turns out to depend on the parameters characterizing the statistical properties of the environment and can be related to the activation and super-activation of memory effects in the open quantum system.

Classical simulations of quantum computing devices generally become intractable as the number of qubits increases. This is due to the exponential growth of the quantum state vector and the associated increase in computational effort. However, when entanglement within the system is limited, rank-adaptive tensor decomposition techniques can be employed to mitigate the exponential scaling. This paper broadens the application of tensor decomposition methods to dynamical simulations of Schrödinger's equation where the Hamiltonian is time-dependent, e.g., to study quantum computing devices subject to time-dependent control pulses. We focus on the tensor-train and Tucker-tensor decompositions that both support low-rank representations, and present an overview of the TDVP, TDVP-2, and BUG, time-integration algorithms for capturing quantum dynamics. The effectiveness of the tensor decomposition approaches is evaluated on representative time-independent and time-dependent Hamiltonian models, with emphasis on how the computational effort scales with the required accuracy and the number of sub-systems in the composite system.

We analyse screening in multi-field scalar-tensor theories, focusing on systems with a dilaton coupled to matter and an axion with a dilaton-dependent kinetic term, in the presence of both planetary and stellar density profiles. Using analytic arguments and fully coupled numerical solutions, we identify a regime in which full screening for a dark-energy-light, effectively unpinned string-dilaton, can occur without fine-tuning. The backreaction of the dilaton's partnered axion field can suppress the exterior scalar charge by selecting a minimum-energy configuration (the BBQ mechanism), yielding robust screening for generic axion gradients. In this regime screening is achieved by cancelling the dilaton's gradient rather than localising it. This reduces the exterior scalar charge and allows for gravity tests in the solar system to be passed. We then show that the more familiar thin-shell intuition need not apply in the multi-field setting. Axion surface gradients can drastically reshape the dilaton profile and drive a more localised transition without generically suppressing the fifth force. The exterior charge can remain essentially unchanged or even be enhanced as the shell is made thinner by a kinetically coupled field. Multi-field two-derivative dynamics therefore decouple localisation in thin shells from screening, evade single-field no-go arguments, and reopen viable parameter space for cosmologically light dilaton-like scalars with strong couplings to matter.

In this paper, we investigate the decay behaviors of three imaginarity-related metrics, specifically the $l_1$-norm-based imaginarity measure, imaginarity robustness, and imaginarity relative entropy, for arbitrary single-qubit pure initial states under three typical quantum channels: dephasing, generalized amplitude damping, and phase-amplitude damping. Furthermore, we extend our analysis to higher-dimensional systems by examining the decay trends of the aforementioned imaginarity metrics for several key two-qubit states under two-qubit channels. We also generalize the concept of the maximal imaginary state (originally defined for single qubits in the resource theory of imaginarity) to separable two-qubit states. In addition, we extend the definitions of imaginary power and de-imaginary power for single-qubit channels to two-qubit channels acting on separable two-qubit states. Finally, we compute the imaginary and de-imaginary powers for several common two-qubit channels.

Recent work has highlighted the deep connection between quantum information and spacetime geometry. Bousso and Shahbazi-Moghaddam (Phys. Rev. Lett. 128, 231301 (2022)) proved that ``hyperentropic'' regions -- where entropy exceeds the area bound -- inevitably lead to singularity formation. In this work, we explore the converse implication: does the thermodynamic consistency of such singularities require them to be hidden? We answer in the affirmative, establishing a Quantum Weak Cosmic Censorship principle governed by Generalized Entropy. This provides a semiclassical mechanism for censorship which forbids naked singularities. Since Quantum Weak Cosmic Censorship is a semiclassical statement, it is more robust than the classical Weak Cosmic Censorship showing naked singularities are forbidden in nature even if quantum effects are taken into account.

We propose collective nuclear polaritons formed by hybridizing a 229Th nuclear ensemble with a vacuum-ultraviolet cavity mode generated via four-wave mixing, achieving a collective light-matter coupling that scales as $\sqrt{N}$. In the strong-coupling regime the system displays vacuum Rabi oscillations, indicating the hybridization between cavity photons and nuclear excitations. In the superradiant regime, the stored excitation is released in a cooperative burst with peak intensity scaling as $N^2$. The emission lifetime shrinks from thousands of seconds to the millisecond scale and remains tunable. Detuning sweeps across the polariton avoided crossing allow adiabatic conversion of the photonic excitation into a collective nuclear excitation, enabling reversible quantum storage. Our results demonstrate that cavity-mediated nuclear polaritons enable deterministic lifetime engineering and coherent quantum storage in nuclear systems.

A quantum channel is defined as being completely positive (CP) and trace preserving (TP). While not every quantum channel is invertible or reversible, every quantum channel admits various kinds of generalized inverses such as the Moore-Penrose inverse and the Drazin inverse. A generalized inverse of a quantum channel may not itself be a quantum channel: it often fails to be CP. However, generalized inverses still play an important role in quantum error mitigation. Here, because it is often desirable for the generalized inverse of a quantum channel to be at least TP, the Drazin inverse, which is TP, is favoured over the Moore-Penrose inverse, which is not in general TP. In this paper, we take a categorical perspective on generalized inverses of quantum channels. This allows us to give a simple proof of the fact that the Drazin inverse of a quantum channel is always TP. It also allows us to show that for unital quantum channels, the Drazin inverse is also unital. We then generalize this result to dagger Drazin inverses, which allows us to show that for unital quantum channels, the Moore-Penrose inverse is both TP and unital as well. This opens the door to new applications of both the Drazin inverse and Moore-Penrose inverse in quantum information theory and, in particular, in quantum error mitigation.

We present a perturbative closed-time-path (in-in) formulation of an optomechanical system in which a quantum field interacts with a moving mirror via radiation pressure. We derive the effective action governing the dynamics of the moving mirror, incorporating the full back-reaction of the cavity field. These effects are encoded in fluctuation and dissipation kernels, that we show satisfy fluctuation-dissipation relations, and whose spectral structure reveals a direct connection with the underlying physical mechanism responsible for the back-reaction, that is particle creation by the dynamical Casimir effect. By deriving the semiclassical equations of motion for the moving mirror, and computing the energy radiated into the field within the in-out formalism of quantum field theory, we verify the energy balance between the mechanical energy dissipated by the optical back-reaction forces acting on the mirror and the energy carried by the particles created in the field.

We report a novel feature of relic gravitational waves (GWs) in non-singular bounce cosmologies that is testable in light of GWs astronomy. In non-singular bounce cosmologies, the effective potential $M_p^2 a^{\prime \prime}/a$ that governs the evolution of primordial GWs contains two peaks due to the existence of contraction phase prior to the standard expansion phase. Accordingly, relic GWs interference between the two peaks, resulting in a distinctive oscillatory feature in the spectrum, analog to the resonant tunneling effect in quantum mechanics. As a result, the GWs spectrum exhibits an oscillatory patterns on high frequecy regime, distinctive to other cosmological scenarios such as inflation. We show that the amplitude of GWs spectrum is high enough to reach the sensitivity of current and forthcoming GWs instruments, making our predictions falsifiable. Hence, our finding offers a promising way to experimentally test the non-singular bounce scenarios and search for new physics in early universe cosmologies.

Quantum criticality has received extensive attention due to its ability to significantly enhance quantum sensing. But its realization and control in many-body quantum systems remain challenging. We present an effective scheme to simulate the Lipkin-Meshkov-Glick (LMG) model using a coherent Ising machine (CIM) composed of a network of degenerate optical parametric oscillators (DOPO). In our work, the spin variables of the LMG model are mapped onto the phases of DOPO pulses, and the spin-spin interactions are realized by all-to-all couplings among them. Through our investigation of the critical behavior in the antiferromagnetically coupled LMG model in the thermodynamic limit, i.e., $N\rightarrow\infty$, and its application in quantum sensing near the critical point, we verify that the CIM does not only effectively capture the second-order quantum phase transition (QPT) at the critical point but also reconstructs its complete phase diagram under ferromagnetic coupling. Furthermore, we demonstrate how the critical dynamics of this simulation platform can be utilized for quantum-enhanced metrology, achieving a measurement precision that diverges near the critical point of the LMG model. These results highlight the capability of the CIM as a flexible experimental platform for investigating the QPT in the fundamental quantum magnetic models, providing valuable insights into quantum simulation and critical phenomena.

The Kirkwood-Dirac (KD) quasiprobability distribution is a fundamental representation for quantum states and has been widely applied in quantum metrology, quantum chaos, weak values in recent years. A quantum state is KD-classical if its KD-quasiprobability distribution forms a valid classical probability distribution with respect to two given bases, and KD-nonclassical otherwise, with the latter being closely associated with quantum advantages in various quantum processes. In this work, we investigate the structural characteristics of the KD-classical state set when the transition matrix between two orthonormal bases takes the form of a discrete Fourier transform (DFT) matrix. First, we adopt an alternative analytical approach to prove that the set of KD-classical states in a $p^r$-dimensional Hilbert space is the convex hull of KD-classical pure states--a conclusion that was recently established by De Bi{è}vre et al [Annales Henri Poincar{é}, 1-20, 2025]. Furthermore, we define a directed graph and use it to characterize KD-classical pure states in a Hilbert space of arbitrary dimension $d$. That is, the convex hull of KD-classical pure states along any path from the start vertex to the end vertex in this directed graph is exactly the intersection of the KD-classical state set and the linear space spanned by these path-associated KD-classical pure states. This general result not only yields the $p^r$-dimensional conclusion in a straightforward manner but also encompasses Theorem 2 in the existing work [J. Phys. A, 57, 435303, 2024], demonstrating its generality and inclusiveness.

Measurement backaction degrades dispersive readout of superconducting qubits even at modest drive strengths, often via the reduction of qubit lifetimes during readout. In this work, we theoretically and experimentally study this degradation and show how it can result from the interplay between detuned two-level systems (TLSs) and a drive-renormalized qubit spectrum. For modest to strong readout, the qubit emission spectrum becomes non-Lorentzian and depends sensitively on the readout drive frequency (even when measurement rate is fixed). We combine the readout-modified qubit emission spectrum with time-dependent perturbation theory to predict qubit lifetimes in the presence of a TLS bath. Master equation simulations and experimental measurements on a frequency-tunable transmon confirm these predictions quantitatively. In particular, we find that driving at the resonator frequency associated with the qubit ground state yields the narrowest qubit emission spectrum and the least lifetime degradation for a fixed measurement rate, providing a practical guideline for optimizing readout protocols in future quantum processors.

Millimeter waves are emerging as an enabling technology for connecting and enhancing different quantum platforms such as Rydberg atoms, optomechanics, and superconducting qubits. In this work, we focus on the interaction between millimeter wave photons and conventional transmon qubits, specifically for qubit readout. We study a circuit quantum electrodynamic (cQED) system consisting of a millimeter-wave cavity at $ω_r = 2π\times 34.7$ GHz and a transmon qubit at $ω_q = 2π\times 3.1$ GHz coupled at rate $g = 2π\times 1.3$ GHz. With such a large detuning between cavity and qubit, $ω_r/ω_q > 10$, we are able to suppress drive induced unwanted state transitions, enabling strong drives for qubit readout. We measure no resonant state transitions up to $1,000$ drive photons and readout the qubit state with more than $100$ photons to achieve a measurement fidelity greater than 99% without the aid of a quantum limited amplifier.

For high-bias operation devices such as silicon carbide (SiC) power devices, early detection of failure mechanisms is essential to ensure reliability. This requires a method to map high electric fields with high spatial resolution, which has not been realized until now. Here we report that the silicon vacancy (Vsi) in SiC has outstanding characteristics for detecting electric fields applied in various directions within a high-biased SiC device. Vsi exhibits an equivalent response to electric field components parallel (Epara) and perpendicular (Eperp) to the c-axis, a feature unique among quantum sensors, and the responsiveness to Epara and Eperp enables detection of arbitrary electric fields encountered in cutting-edge SiC power devices. We confirmed high electric field detection of ~2.3 MV/cm, which is ~90% of the breakdown electric field of a 4H-SiC with typical carrier concentration. Selectively formed Vsi enables high-resolution mapping of electric field distribution. Vsi-based quantum sensors bring data-driven research and development methodologies as well as device degradation diagnosis.

We investigate the Magnus expansion for a generic time-dependent two-level system under single-axis driving. By virtue of the \(\mathfrak{su}(2)\) Lie algebra, the expansion is decomposed into a commutator-free form. To illustrate the usefulness of the gained expression, we then revisit the Landau-Zener-Stückelberg-Majorana model, with a focus on non-adiabatic transitions as well as the Stokes phase. In addition, the semiclassical Rabi model is systematically treated by determining the Floquet quasienergy up to different orders. We demonstrate how to employ suitable picture transformations as well as on how to enforce the symmetry of the underlying model in order to guarantee convergence of the expansion as well as to achieve satisfactory agreement with the exact results. For both models that we studied it turns out that a third order approximation yields results that are in next to perfect agreement with exact analytical ones. Surprisingly, in the case of the semiclassical Rabi model, even the second order Magnus approximation in the adiabatic picture produces almost exact results over the whole parameter range.

In this study, we developed a diamond quantum magnetometer based on Ramsey interferometry with a short sensor-to-sample distance. Conventional biomagnetic sensors with ensemble nitrogen-vacancy centers using continuous-wave optically detected magnetic resonance and Ramsey methods typically rely on watt-level lasers to achieve high sensitivity, resulting in thermal issues. In contrast, by employing the light-trapping diamond waveguide technique in a high-pressure and high-temperature diamond sample treated with electron beam irradiation, we obtained a high photon conversion efficiency of 9.5%, enabling us to simultaneously achieve a high sensitivity of 2.93(7) pT/Hz^1/2 in the 100-400 Hz frequency range and a minimal temperature increase of only approximately 13 K at a low laser power of 210 mW. Using a dry phantom designed to mimic magnetoencephalography signals, we measured a weak magnetic field of 77.7(2) pT without signal averaging at a sensor-to-sample distance of 2.5 mm. This short-distance measurement prevents severe spatial signal attenuation, yielding a high signal-to-noise ratio. The development here is crucial for practical biomagnetic applications based on Ramsey interferometry.

In measurement-based quantum computation (MBQC), quantum circuits are implemented using adaptive measurements on an entangled resource state. In practice, the resource state will always be prepared with some noise, and it is crucial to understand the effect of this noise on the operation of MBQC. Typically, one measures the fidelity of the noisy resource state with the assumption that a high fidelity state means a high fidelity computation. However, the precise relationship between these two fidelities is not known. Here, we derive an expression that equates the average fidelity of the MBQC output state to a certain correlation function evaluated on the noisy resource state. Using this expression, we show that state fidelity provides a tight lower bound on average MBQC fidelity. Conversely, we also find that state fidelity can greatly underestimate average MBQC fidelity, implying that state fidelity is not a good indicator of MBQC performance in general. In response, we formulate an efficient method to directly estimate average MBQC fidelity by measuring the aforementioned correlation function. These results therefore improve our ability to characterize noisy resource states in quantum computers and benchmark MBQC performance.

Quantum channels generally reduce the distinguishability of quantum states, limiting information transmission and processing. Previous work introduced a protocol capable of increasing the distinguishability of states after the action of a specific quantum channel. Here we show how to systematically determine the maximal distinguishability gain achievable by this method. We develop an algorithm that identifies the optimal implementation of the protocol and applies to arbitrary quantum channels in a straightforward manner. Using this approach, we demonstrate that a weakly non-Markovian channel can effectively be converted into an essentially non-Markovian one through a distillation-like process. We further analyze the quantitative features of the optimized protocol, characterizing the conditions under which the enhancement is most pronounced. Our results provide a general framework to assess and optimize distinguishability recovery in open quantum systems.

Modern cryptography relies on keyed symmetric ciphers to ensure the secrecy and authenticity of high bandwidth data transfer. While the advent of quantum computers poses a challenge for public key cryptography, unbroken ciphers are considered safe against quantum attacks if their key is sufficiently long. However, concrete bounds on the required key length thus far remain elusive: Despite the well known asymptotic complexity of Grover's quantum search, the optimal algorithm to recover a secret key, no implementation-agnostic tight bounds were established. Here, we discuss the quantum thermodynamic limits of generic search algorithms, and find a work-runtime trade-off for autonomous computers with a fundamental lower bound. By devising an application-specific quantum protocol, which outperforms circuit and adiabatic implementations, and saturates this bound, we demonstrate that it is tight. Applying this limit, we find that a secret key of 831 bit length cannot be reconstructed deterministically in an expanding, dark-energy-dominated universe until star formation is expected to cease. Implications for post quantum cryptography, and quantum key distribution are discussed.

Accurate control of quantum degrees of freedom is promising for sensing, communication, and computing, but building a useful quantum computer faces a central isolation-and-control challenge: qubits must remain well isolated from their environment to preserve coherence, yet also be coupled strongly enough for control, readout, and reset. Existing approaches address this challenge only partially, using separate reset elements, drive schemes, and Purcell filters, often with added complexity and tradeoffs such as heating and crosstalk. Here we introduce and demonstrate a first-generation quantum dial: a device that on demand mediates the coupling of a qubit to selected auxiliary degrees of freedom. Our implementation uses a band-stop filter between a high-coherence transmon qubit and a broadband transmission line, enabling the coupling strength to be tuned by several orders of magnitude on nanosecond timescales without significant Stark shift. In the reset configuration, we reduce the qubit energy relaxation time T1 from >150 $μ$s to about 200 ns and demonstrate reset independent of the initial state. In the control configuration, we obtain 99.99% idle fidelity and 99.9% gate fidelities for 40 ns pulses at about -110 dBm. The same device also enables thermometry of the qubit environment, reaching a noise-equivalent temperature of 0.6 mK/$\sqrt{Hz}$ at 60 mK and approaching the Cramér-Rao bound at higher temperatures. The quantum dial thus offers fast, on-demand switching between isolation and strong coupling, with potential to reduce noise and errors in future quantum processors.

Wu et al. introduced residual quantum correlations (RQC) in 2015 and defined them in terms of two complementary bases. Given a measure for classical correlations, its optimization defines a local basis. Relative to this local basis, one defines a new one that is mutually unbiased to the first one. In the latter, the corresponding measure for quantum correlations is calculated. Local available quantum correlations (LAQC) define a measure for maximal RQC and were introduced by Mundarain and Ladron de Guevara. In previous articles, we derived an analytical exact solution for this measure for 2-qubit X states. Using those results and deriving an expression for the RQC measure introduced by Wu et al., we analyze their behavior for two non-Markovian quantum dephasing channels: Random Telegraph (RT) and Modified Ornstein-Uhlenbeck (MOU) noises. We derive general conditions for sudden death and revival of RQC in X states and illustrate these results with three families of bipartite qubit states: Werner states, Maximally Nonlocal Mixed States (MNMS), and Maximally Entangled Mixed States (MEMS).

Pulsar Timing Array (PTA) experiments have entered a new era with evidence for a nanoHertz gravitational wave background (GWB). This review describes the physics of detection, detailing the noise models and cross-correlation techniques required to isolate the Hellings-Downs curve. We discuss astrophysical implications, arguing that the perceived tension between current amplitudes and standard merger models is largely resolved by new insights into supermassive black hole binary populations. Beyond the stochastic background, we review the framework for multi-messenger continuous gravitational-wave searches, highlighting targeted search campaigns and rigorous detection protocols. We also examine the potential to probe New Physics, including cosmic strings and ultralight dark matter. Critical challenges are addressed, including small-scale leakage bias in anisotropy searches and the separation of deterministic signals from the GWB and pulsar noise. Finally, we outline the field's future, from rapid data combination strategies to the sensitivity gains expected from the Square Kilometre Array Observatory (SKAO) and DSA-2000.

Extreme Mass Ratio Inspirals (EMRIs) constitute a prime target for future space-based gravitational-wave observatories such as LISA. In this paper, we analytically investigate the long-term phase shift (dephasing) in the gravitational wave signal induced by a first-order quantum chromodynamics (QCD) phase transition within a neutron star orbiting a supermassive Kerr black hole. By modeling the transition from a hadronic phase to a quark core phase, we quantify the sudden change in the tidal deformability ($Λ$) of the secondary object. Utilizing the Teukolsky formalism and Post-Newtonian expansions, we derive a strict analytical scaling law for the accumulated dephasing. We demonstrate that the Kerr spin parameter $a$ and the critical phase transition orbital velocity $v_c$ significantly amplify the dephasing effect. Our analytical framework provides a robust tool for probing the non-perturbative QCD equation of state at high baryon densities using gravitational wave astronomy.

Bell theorems of many-body nonlocality and contextuality serve as a benchmark for proving quantum advantage in that a quantum computer outperforms a classical computer for a certain problem. In practice, however, near-term quantum devices do not prepare perfectly pure states but rather mixed states produced from noisy channels. We investigate noisy quantum advantage by considering thermal mixed states of one-dimensional many-body systems with a symmetry-protected topological (SPT) order. In the pure-state (or zero-temperature) case, these states are known to be useful for measurement-based quantum computation, and to outperform classical computers in a many-body contextuality game, provided string order parameters (SOPs) of SPT are sufficiently large. Here, we show that quantum advantage in mixed states is measured by a combination of twisted SOP and symmetry representation expectation values. Using the minimally entangled typical thermal states algorithm, it is demonstrated that quantum advantage persists to a nonzero critical temperature for finite-sized instances of the many-body contextuality game. While this critical temperature goes to zero in the thermodynamic limit, it is relatively robust to system size, suggesting that these states remain useful for demonstrating genuine "quantumness" of noisy hardware in a scalable fashion. Finally, we show that the quantum winning probability is lower bounded by the global fidelity with the 1D cluster state, so that our contextuality game can provide an operational meaning to benchmark the capacity to create long-range order like SPT states in near-term experimental devices.

In this work we find solutions of the ($n+2$)-dimensional Einstein Field Equations (EFE) with $n$ commuting Killing vectors in vacuum. In the presence of $n$ Killing vectors, the EFE can be separated into blocks of equations. The main part can be summarized in the chiral equation $\ (αg_{, \bar{z}} g^{-1})_{, z} + \ (αg_{, z} g^{-1})_{, \bar{z}} = 0$ with $ g\in SL(n,\mathbb{R})$. The other block reduces to the differential equation $(\ln f α^{1-1/n})_{, z} = 1/2 αtr( g_{, z} g^{-1})^2$ and its complex conjugate. We use the ansatz $g = g(ξ) $, where $ξ$ satisfies a generalized Laplace equation, so the chiral equation reduces to a matrix equation that can be solved using algebraic methods, turning the problem of obtaining exact solutions for these complicated differential equations into an algebraic problem. The different EFE solutions can be chosen with desired physical properties in a simple way.

The development of magnetoelectric materials requires chemical design strategies that integrate structural polarity with magnetic lattices capable of supporting competing spin interactions. Here, we demonstrate such an approach in the polar, buckled honeycomb magnet Co2SeO3Cl2. Magnetization and heat-capacity measurements reveal strong magnetic anisotropy and four successive magnetic transitions at 25.4, 16.8, 11, and 3 K. The recovered magnetic entropy through the ordering regime is only around half of the expected 2Rln(2), indicating persistent spin fluctuations. Second-harmonic generation measurements show three pronounced intensity anomalies at 11, 17, and 26 K that coincide with magnetic transitions while revealing that the crystallographic symmetry is preserved. Together, these results demonstrate that polar, buckled honeycomb magnets offer an unconventional phase space for coupling magnetic and electric dipoles in magnetoelectric materials.