arXiv Daily Digest - 2026-06-15

Verifier-driven self-DPO is a common recipe for self-improving production visual-language models. In this setup, a frozen verifier scores candidate generations, the top- and bottom-scoring candidates form a preference example, and DPO updates the learner. The deployment-time assumption is monotone: a stronger verifier should yield a stronger student. We show that this assumption can fail because verifier quality is highly task-specific. On a four-rung open-source verifier ladder across MathVista, MMMU, and BLINK, the same verifiers that are above-threshold and improve a Qwen-3-VL-2B student on MathVista become sub-threshold on MMMU, where their task-rubric accuracy drops to 8% to 23%. In this regime, every verifier we tested silently regresses the student, producing drops of 3.4 to 10.9 percentage points below the frozen baseline while the DPO training loss continues to decrease. The regression replicates on a second student, Qwen-2.5-VL-3B. Moreover, within the failure regime, damage is confidence-inverted: the more accurate-but-still-wrong verifier causes larger regression than a near-random verifier, suggesting that progress-gated replay amplifies confidently wrong preference pairs. We give a compact mechanistic explanation via a variance theorem for progress-gated replay and its direction-mismatch failure mode. The deployment message is operational rather than purely diagnostic: before running any verifier-driven loop, teams should measure target-task rubric accuracy, rank verifiers by target-task rubric quality rather than parameter count, and treat diminishing returns in above-threshold regimes as a verifier-side compute budget cap.

The global use of artificial intelligence has increased interest in assessing the ability to generate culturally localized content, including stories. Cultural localization in stories often occurs through either templated localization -- the use of cultural markers (e.g., names, locations) in a generic narrative -- or holistic localization -- the variation of plots, values, and themes, in addition to cultural markers. We propose a method to measure the degree to which content was generated through templated localization. Specifically, we identify the lexical tokens that distinguish stories across nationalities and measure the similarity of the narratives that remain after removing them. In stories generated by five models on 125 topics for 193 nationalities, our method is able to detect that only a small subset (9-17%) of the vocabulary accounts for the variation across nationalities and that the narratives that remain after removing them contain repeated multi-word sequences, suggesting the presence of a shared culturally-agnostic narrative template. Finally, we characterize the cultural markers for their stereotypicality and offensiveness, finding that markers from 19 countries, mostly located in the Global South, are on average offensive.

Open diffusion language models are marketed as parallel, non-autoregressive decoders, yet the order in which a shipped checkpoint actually commits its tokens is almost never measured. We instrument DiffusionGemma 26B, a masked discrete-diffusion mixture-of-experts model built on Gemma 4, hooking its sampler's accept step to record which canvas positions commit, when, and at what confidence. Across a 686-prompt, six-regime probe suite we find that its decoding is neither parallel nor block-autoregressive: it follows a partial left-to-right commit bias whose apparent strength depends almost entirely on the granularity at which you look. Order is weak token by token and strengthens smoothly as the analysis is coarsened, so the model's "block size" turns out to be an artifact of the measuring ruler rather than the architecture. The model commits in large simultaneous batches, leaving much of the within-batch order genuinely undefined rather than merely unobserved. The behaviour is regime-dependent: structured JSON is committed in essentially arbitrary order, and a position's commit confidence tracks correctness on mathematical reasoning but carries no signal on factual recall. Commitment is aggressive, finishing in a short late burst well inside the step budget, while task accuracy matches the model's autoregressive Gemma-4 sibling. Beyond these findings, our central contribution is methodological: measuring decoding order honestly demands handling trailing-EOS padding, within-regime confounding, commit non-monotonicity, block-size sensitivity, and large commit-batch ties, each of which can otherwise manufacture a decoding-order result that is not really there.

Machine learning (ML) and Artificial Intelligence (AI) components are increasingly embedded in software products, yet undergraduate software engineering (SE) curricula rarely provide systematic preparation for building, testing, deploying, and maintaining AI/ML-based software systems. This paper aims to provide evidence-based guidance for integrating AI/MLrelevant content into core SE education. We compile and define a structured inventory of topics relevant to SE practice in AI/MLbased software, then map these topics against required courses in a set of representative SE curricula to identify coverage gaps. To assess educational priorities and feasibility, we survey SE instructors on topic importance and integration constraints. Based on the crosswalk between topic definitions, curriculum coverage, and instructor prioritization, we derive a guideline that recommends where and how high-priority topics can be embedded within existing SE courses.

Over time, GitHub has introduced different strategies for sharing reusable code artifacts. In addition to fork-based reuse, template repositories provide a distinct feature for generating new projects from scaffolding. Although this feature has been available since 2019, little is known about the domains it supports, its maintenance characteristics, or the practices that guide practitioners for effective template design. To address this gap, we conduct a large-scale empirical study of GitHub template repositories across the five most used programming languages. First, we mine and categorize templates to analyze the domains they serve, exploring the LLM-as-a-judge strategy. Next, we explore the reliability of templates by evaluating the associations between repository characteristics and activity, and quality-related issues (e.g., code smells, vulnerabilities, and security hotspots) through statistical analysis. Finally, we qualitatively analyze a representative subset of templates to derive practical guidelines and recurring pitfalls for template design and management. Our results show that Web Development is the predominant domain across ecosystems, while maintenance and quality issues vary by programming language. We further find that high-quality templates tend to adopt established software engineering practices, while providing comprehensive documentation and clear guidance for use. Overall, our findings offer empirical insights and actionable guidance to support practitioners in designing and adopting high-quality template repositories.

We show that the three movements of Beethoven's "Moonlight Sonata" (Op. 27 No. 2) instantiate three distinct machine learning architectures -- not by analogy, but by structural correspondence. Through computational analysis of the score (entropy, Jensen-Shannon divergence, dissonance, hand distributional overlap, self-similarity matrices, temporal memory decay, and contextual pitch embeddings), we establish four counterintuitive findings: (1) perceived musical "temperature" is governed by throughput, not distributional width; (2) the lightest movement carries the highest dissonance; (3) the movements implement streaming, recurrent, and periodic positional encoding memory architectures; and (4) the same pitch class acquires different contextual identities across movements, analogous to contextual vs.static embeddings in NLP -- and unsupervised clustering recovers the tonal structure without music-theoretic input. We construct a reverse sonification (decoding analytical features back into MIDI) and quantify the chirality of the encode-decode cycle: what distributions preserve and sequential ordering destroys. Prompted by a listener's observation that the decoded piece sounds like "mirror isomers that can't be superimposed," the chirality measurement reveals reconstruction loss increasing monotonically with n-gram order. Bootstrap baselines and subsample checks confirm all movements carry sequential information above noise, though raw values are confounded by sample size. Cross-domain comparison shows natural language has higher chirality than music, reflecting stronger sequential constraints.

Survival prediction plays a central role for healthcare providers and clinical researchers. Accurate risk stratification enables early intervention and improved patient management. Most existing deep survival models learn one common feature representation for all patients, which may hide important differences between patient subgroups. In contrast, a Mixture-of-Experts (MoE) framework allows different parts of the model to focus on different patient patterns, leading to more individualized representations. Therefore, in this work, we propose a mixture-of-experts enhanced adaptive deep clustering survival framework (AdaCSM) for modeling such heterogeneous survival patterns. We introduce a routing-based expert mechanism that enables conditional specialization within a parametric survival modeling framework. The proposed architecture allocates patients to specialized risk predictors dynamically while preserving the patient survival and subtype clustering objectives. We compare our method with state-of-the-art survival and deep clustering models on multiple real-world longitudinal clinical cohorts spanning diverse disease domains. The proposed method demonstrates improved predictive performance and leads to interpretable results in survival analysis.

Concurrent programming is a core component of Computer Science curricula, yet remains notoriously difficult for students to master due to its inherent complexity and the nondeterministic nature of concurrency bugs such as deadlocks and race conditions. In this work, we present ParaView, an educational tool designed to help students understand, debug, and correct concurrency issues in parallel programs written in C/C++. ParaView provides transparent execution recording and visualization to make parallel execution observable and comprehensible. We evaluated ParaView through a series of debugging and implementation tasks, with 17 students participating. Results showed a significant improvement in debugging and implementation successes compared to previous course iterations. A student survey confirmed that most participants found ParaView helpful. To further support learning outside the classroom, we explored using Large Language Models (LLMs) to analyze concurrency bugs and suggest fixes. While LLMs were highly effective in identifying bugs and explaining execution traces, the correctness of their bug fixes varied, especially for more complex synchronization patterns. Our findings suggest that recording-visualization tools like ParaView, complemented by artificial intelligence (AI), can improve teaching and learning of concurrent programming.

Wearable devices and smartphones generate rich behavioural time series that can support proactive health interventions, yet systematic comparisons of modern forecasting architectures for these data are lacking. In particular, it remains unclear how models generalise across populations, how different architectures respond to participant-level fine-tuning and how forecasting accuracy degrades across multi-day horizons. We benchmark six deep learning architectures, two zero-shot Foundation Models (FM) and statistical baselines on three public datasets encompassing over 800 participants, reporting per-feature metrics for step counts, screen time and sleep duration across 1-8 day horizons. We further conduct a per-feature personalisation study across all six architectures and assess FM transferability across dataset sizes and temporal granularities. Our key findings are: (i) no single architecture dominates, PatchTST leads among trained models while the three runners-up (TCN, MLP, Transformer) show no meaningful performance difference; (ii) the FM TimesFM matches or exceeds trained models zero-shot, especially in low-data regimes and (iii) participant-level fine-tuning reduces per-feature RMSE by 16-60\%, with sleep benefiting most and step counts least. These results provide practical guidance on architecture selection, FM applicability and personalisation strategies for mobile health forecasting. To the best of our knowledge, this is the first study to jointly evaluate modern deep learning, FMs and personalisation for multi-horizon behavioural forecasting from wearables.

This study presents a statistical and machine learning framework for characterizing a hydrogen-based multi-energy system (H-MES) using one year of high-resolution operational data. Statistical analysis revealed a binary operation driven by renewable surplus, with solar irradiance explaining 45.7% of rank-based variance in hydrogen production, a large effect by conventional standards. Only high-irradiance periods triggered meaningful electrolyzer engagement, while electricity demand exerted a weaker inverse suppression effect ($ε^2 = 0.126$). Multiple regression confirmed electrolyzer power as the dominant linear predictor, with a synergistic solar-wind interaction. Notably, Random Forest analysis ranked wind output first in predictive importance despite its weak bivariate correlation (r = 0.167), revealing non-linear dynamics invisible to parametric methods. A sequence model exploited strong 24-hour autocorrelation (r = 0.845) for operational forecasting, while a reinforcement learning agent optimized hydrogen revenue dispatch. The core contribution is demonstrating that statistical and machine learning approaches are complementary for H-MES modeling and control.

Online group chats are social spaces with local conversational norms that are rarely stated explicitly. The ability and willingness of LLM-based agents to recognize and adapt to these norms remains mostly unexplored. We introduce LoSoNA, a benchmark for local social norm adaptation in multi-party chat. Each scenario gives a subject model a curated group-chat transcript in which non-subject participants demonstrate a hidden local norm, followed by a final elicitor turn that forces a response revealing whether the subject has inferred that norm. We evaluate eight frontier and open-weight models under four prompting conditions that vary how explicitly the model is told to treat the prior conversation as evidence for how it should answer. Naive prompting remains limited for most models; explicit norm-aware prompting helps unevenly, with Gemini 3.1 Pro reaching $84.2\%$ and Claude Fable 5 reaching $81.6\%$, while several other models show small gains or regressions. LoSoNA contributes to recent calls for evaluating LLM social capabilities by testing whether models can infer local conversational norms from precedent and use them in a one-turn group-chat response.

Post-training INT8 (W8A8) quantization of diffusion transformers is widely deployed as a speed optimization, yet on consumer Ampere GPUs it is frequently slower than the FP8 and NF4 alternatives it is meant to beat. We trace this to a software artifact: the production "INT8" forward quantizes weights and activations only to immediately dequantize them back to bf16 and run a bf16 matrix multiply, never engaging the GPU's INT8 tensor cores, so the hardware's compute advantage is left entirely unrealized. We close this gap with a single fused Triton INT8 GEMM (int8xint8->int32 on Ampere tensor cores, with per-token x per-channel dequantization and bias folded into the epilogue, autotuned per GEMM shape) dropped into the Ideogram 4.0 diffusion transformer's linear layers in place of the dequantize-to-bf16 path. In the kernel, the int8xint8->int32 accumulation is bit-exact against torch._int_mm and the dequantized output matches the reference at cosine similarity 1.0 with no NaNs, running 2.8-4.2x faster than bf16 per GEMM. End to end it delivers a ~1.1x (~9-10%) speedup at 768px, and at 1024px it generates an image in 156.5 s on a single RTX 3090, faster than the single-card NF4 (164.5 s) and FP8 (172.9 s) baselines, at no measurable quality cost on these point estimates (PickScore/CLIPScore). INT8 thus goes from the slowest variant to the fastest, and 1024px becomes single-GPU feasible. The primary speed criterion (beat FP8, by ~9.5%) is comfortably met; the NF4 margin (~4.9%, single-run n=4) is within run-to-run variance we did not quantify and is best read as consistent with meeting the stretch target. We close with an honest deployment map: the win is specific to consumer Ampere, and on A100 and B200 the same kernel loses to those cards' fast native bf16/FP8 paths.

Transformer-based neural operators have shown remarkable performance for approximating solution operators of partial differential equations on complex geometries. However, existing approaches implicitly assume a fixed domain size, which limits their ability to generalize at inference. In this work, we investigate domain extension, namely zero-shot inference on spatial domains that are significantly larger than those encountered during training. We argue that this setting fundamentally requires spatial locality and translation equivariance. We propose to implement this locality via a decomposable bias in the attention logits computation, enabling finely controllable locality while remaining fully decomposable into query-key inner products and directly compatible with optimized attention kernels. Combined with rotary positional embeddings, it enables expressive embeddings with controllable spatial support without altering the transformer architecture. We empirically show that our approach substantially improves zero-shot generalization to larger domains across two PDE benchmarks and a 3D industrial atmospheric flow application. Our code and datasets are available at https://github.com/cerea-daml/domain-extension.

AI-assisted software development has moved from line-level autocomplete to agents that can plan changes, edit files, and submit pull requests with limited human supervision. Open-source software, however, evolves through a process designed for humans: contributor agreements, codes of conduct, and review norms all assume a legally accountable person who can attest to provenance and answer reviewer questions. Autonomous and semi-autonomous AI contributors strain those assumptions, and the 2025-2026 record of agent-driven incidents, AI-generated nuisance volume, and platform-level shutdowns shows that the gap is operationally consequential. Several open-source organisations have responded with contribution policies, but the result is fragmented, and its alignment with emerging AI governance frameworks (EU AI Act, NIST AI RMF with the UC Berkeley Agentic AI Profile, ISO/IEC 42001 and 23894) is unmapped at the contribution level. We compare policies across six organisations (SymPy, LLVM, matplotlib, OpenInfra, the Apache Software Foundation, and the Linux Foundation) using Most-Similar Systems Design with indicator-based coding and process tracing for SymPy and LLVM. From this we derive a six-dimensional taxonomy (disclosure, responsibility, human oversight, licensing, enforcement, maintainer workload), an ordinal Policy Maturity Score, and a mapping of documented agent incidents onto the dimensions each policy fails to govern. Aligning the dimensions with the regulatory frameworks above identifies overlapping gaps neither side currently closes, and we close by sketching the shape of a harmonised tiered framework and the empirical evaluation needed to calibrate it.

Clustering is widely used for exploratory analysis and scientific discovery, driving insights from market segmentation to biological data analysis, but its outputs can be difficult to interpret, audit, and reproduce as modern datasets become increasingly large and complex. Reliable use of clustering requires understanding which features drive the discovered structure, yet feature-level explanations for clustering remain scarce compared with methods in supervised learning. Furthermore, existing clustering feature importance scores are often tied to specific algorithms and data assumptions. To address these challenges, we propose Cluster LOCO (Leave-One-Covariate-Out), a family of model-agnostic feature importance scores for clustering. Cluster LOCO is built on feature occlusion and clustering generalizability, defined as whether cluster labels learned on one subset of the data can be accurately predicted on held-out samples. For any chosen clustering algorithm, Cluster LOCO quantifies a feature's importance by measuring how much its removal degrades generalizability. We first introduce Cluster LOCO-Split, which relies on data splitting, and then extend it to Cluster LOCO-MP, a minipatch ensemble-based version designed for large-scale data. Across synthetic simulations and an application to cell-type discovery in single-cell transcriptomics, we show that Cluster LOCO more reliably recovers informative features than existing clustering feature importance methods.

Large Audio-Language Models (LALMs) have shown strong performance on a wide range of audio understanding tasks, yet they still struggle with complex audio reasoning. A practical way to improve such capabilities is post-training, whose effectiveness critically depends on the quality and diversity of training data. However, existing audio-language datasets often contain substantial redundancy, where many samples are highly similar in acoustic content and thus provide overlapping supervisory signals. Such redundancy not only increases annotation cost, but also limits corpus diversity and reduces the effectiveness of post-training. To address this issue, we propose a redundancy-aware data construction pipeline for building reasoning-oriented supervision for LALMs. Specifically, we first perform acoustic similarity-based deduplication across raw audio datasets to improve corpus diversity. We then integrate existing audio captions and question-answer pairs into a unified multiple-choice format. Based on these unified annotations, we leverage Qwen3-30B to generate chain-of-thought (CoT) rationales for reasoning-oriented supervision. Based on this pipeline, we construct AudioDER, a reasoning-oriented post-training dataset containing approximately 191k samples spanning sound, speech, and music. Each sample consists of an audio clip, a multiple-choice question, four answer candidates, an audio caption, and a CoT rationale. Extensive experiments show that post-training on AudioDER consistently improves the performance of Qwen2-Audio-7B-Instruct on multiple audio reasoning benchmarks, including MMAU-mini, MMSU, and MMAR. We hope AudioDER can serve as a valuable resource for advancing audio reasoning research and the development of more capable LALMs.

LLM agent systems increasingly run as long-lived autonomous runtimes: scheduling jobs, calling tools, maintaining memory, and pushing results to humans. We present a longitudinal study of silent failures in one such system: a personal-assistant agent runtime in continuous production since March 2026, with roughly 40 scheduled jobs, 8 LLM providers, a tool-governance proxy, and a knowledge-base memory plane, defended by 4,286 unit tests and 827 governance checks. Over eight weeks we documented 22 incidents with full root-cause postmortems, in which one meta-pattern -- a failure whose error signal never reaches a human in actionable form -- manifested at least 28 times. We derive a five-class, mechanism-oriented taxonomy: (A) environment and platform quirks, (B) design-assumption mismatches, (C) error swallowing and dilution, (D) chained hallucination and fabrication, (E) operational omission and forensic blind spots. Class D is unique to LLM systems and the most dangerous: the system does not merely fail to report an error -- the LLM transforms it into fluent, plausible narrative delivered to the user. We term this fail-plausible: gray failure's differential observability escalated -- the observer is not just blind, it is convincingly lied to by the failure itself. Three findings: about 70% of silent failures were caught by human user-view observation, not tests or audits; a retrospective audit of 15 incidents found 0% ex-ante prevention but 87% regression blocking -- audits are regression engines, not prediction engines; incident latency (13 hours to 60 days) tracks failure mechanism, not code complexity -- the longest-lived failures lived in the seams between components, where no test runs. We describe the resulting defense framework and distill design principles for agent systems whose failures are loud, attributable, and boring. All postmortems and artifacts are public.

Generative dynamics models enable planning in challenging robotic systems, but safe deployment requires reliably detecting policy-induced out-of-distribution (OOD) transitions. Existing methods typically treat the learned dynamics as fixed and attach post hoc support surrogates. We show that these surrogates can fail when the dynamics are locally insensitive to critical action choices: unsupported control actions may produce latent predictions that resemble demonstrated transitions, suppressing OOD signals despite large true predictive errors. To address this, we introduce support-conditioned control-sensitivity regularization, which promotes sensitive local response to control input changes in learned dynamics in high-support training regions. This preserves control-induced variation while limiting unstable extrapolation due to weak empirical support. Experiments in vision-based obstacle avoidance, manipulation, and real-robot navigation show improved OOD detection and safer closed-loop planning.

Efficient route optimization play a vital role in ensuring both safety and punctuality in railway operations. It is very crucial particularly in heterogeneous multi-gauge railway networks with varying train speed, stopping pattern, infrastructure compatibility constraints increase coordination complexity. In single-track systems these challenges are further intensify due to all trains to share the same track and requires frequent track switching.Stochastic disruptions events including blocked tracks, blocked trains, engine failure and speed slowdowns introduces additional unpredictability in operations and deviate the timetable. However, existing studies predominantly focuses on high-level timetabling, omitting operational details such as track switching coordination. As a result leaving decision to human operators, increasing safety risks into railway operations. This study proposes a framework based on temporal planning for dynamic route optimization and disruption management in heterogeneous railway systems. The framework formulates railway operations as a temporal planning problem using PDDL 2.1 with explicitly modeling gauge compatibility constraints and diverse disruption scenarios. It generates conflict-free timestamped operational plans specifying both optimized schedules and executable action sequences. To evaluate the proposed framework, we developed a benchmark problem set with 200 instances using up to 1,000 track points and 120 trains. Two state-of-the-art temporal planners and a plan validator were employed to assessed the framework. The experimental results demonstrate that the framework effectively generates temporal operational plans for heterogeneous railway systems and handles multi-gauge constraints, disruptions, and reduces dependence on manual decision making.

Granting LLMs direct control over costly, irreversible scientific experiments leads to unsafe exploration and unstable performance, but discarding LLM creativity entirely sacrifices significant optimization potential. We introduce CARE (Controlling LLM-Generated Policies through Auditable Review of Evidence in Scientific Experimentation), an auditable controller for high-throughput experimentation (HTE) optimization that keeps a non-LLM incumbent optimizer as the default action path while using LLMs to revise challenger ranking policies. Before each outcome is revealed, a public-evidence intervention gate compares the challenger with the incumbent. It authorizes the challenger's selection only when the evidence available before selection supports the change, with the decision recorded in the audit log. CARE outperforms all other evaluated methods on Minerva/Olympus and ChemLex benchmarks, with final-best improving from 80.0 to 88.5 on Minerva/Olympus and from 83.9 to 92.1 on ChemLex, relative to the public incumbent. Our experiments indicate that LLM self-evolution is more reliable when it expands the proposal space under an auditable controller, rather than directly choosing experiments.

Identifying persuasive rhetorical cues is critical across domains, from detecting information manipulation and improving AI safety to advancing public health communication. We propose Persuasion Index (PI), a taxonomy of 15 dimensions grounded in persuasion theories from psychology and communication, and one transparent implementation using 55 sub-features built from lexicons and rule-based detectors. The taxonomy is modular: individual detectors can be replaced while preserving the theoretical structure. By evaluating PI on four public datasets varying in domain, style, and outcome measures, we show that PI provides a shared feature space for interpreting rhetorical patterns associated with persuasion-related outcomes. Linear models show that PI features carry meaningful predictive signal while remaining computationally lightweight. Dimension-level analyses reveal recurring associations between PI dimensions and persuasion outcomes across datasets, while also highlighting topic- and stance-specific variation. We release PI as an open-source package and web interface for principled and auditable analysis of human and AI-mediated communication.

When applying Group Relative Policy Optimization (GRPO) for GUI Grounding, rollouts are sampled from a single screenshot view; groups often become either all failures on difficult instances or all successes on easy ones, yielding no useful relative advantage. We propose VISTA (View-Consistent Self-Verified Training), a GRPO-based training framework that constructs each comparison group from multiple target-preserving views of the same GUI instance.Each view is generated by a crop that keeps the target element visible and remaps its box exactly, so model rollouts are compared across semantically equivalent but geometrically different inputs. To stabilize short coordinate generation without turning reinforcement learning into unconditional imitation, VISTA further adds a self-verified cross-view anchor: an oracle answer optimized with an advantage-weighted loss, excluded from the group baseline and activated only when the model has produced a maximum-reward rollout. Across five GUI-grounding benchmarks and multiple Qwen backbones, VISTA consistently improves grounding accuracy.On ScreenSpot-Pro, it raises Qwen3-VL 4B/8B/30B-A3B from 55.5/52.7/53.7 to 63.4/65.8/67.0. Robustness analyses further show higher worst-view accuracy and lower prediction flip rates.

Large language models (LLMs) are increasingly deployed as planners for autonomous agents in household environments. While existing benchmarks evaluate whether LLM-generated plans execute successfully, they overlook a critical type of failure: latent failures. Unlike immediate failures that trigger instant feedback at execution time and enable timely correction, latent failures do not immediately halt plan execution but silently compromise goal achievement. In severe cases, they cause irreversible harm. To address this gap, we introduce SIMMER, a benchmark for evaluating latent failures in LLM planning through a human-curated symbolic world model grounded in the kitchen domain. SIMMER defines a world model comprising 77 actions, 262 unique objects, and approximately 46,800 possible interactions that are semantically realistic, derived from real-world cooking scripts. It then leverages a state machine executor that validates plans against the world model and detects immediate precondition violations, latent hazards, and irreversible failures. Experiments across six LLMs show that even frontier models achieve at most 17% error-free plans. Moreover, up to 56% of plans contain latent failures, the majority of which lead to irreversible consequences. We further demonstrate that explicit state reasoning via counterfactual foresight simulation can reduce latent failures by up to 72% and irreversible cases by up to 75%, suggesting a promising direction for more robust LLM planners.

A central role of personal-agent memory is to turn stored information and prior interactions into future-oriented assistance. In daily use, useful cues come from what the agent observes and how the user interacts with the agent, and the agent must carry them forward from the current request to similar future tasks. Existing memory benchmarks usually test dialogue recall or task improvement in isolation, leaving the trajectory from streaming observations to later assistance largely untested. We introduce StreamMemBench, a streaming benchmark that constructs a two-step task sequence around each evidence anchor from EgoLife egocentric streams. The initial task tests evidence use, while the follow-up task tests whether feedback and interaction experience are reused. Four metrics diagnose evidence recall, initial evidence use, feedback incorporation, and follow-up reuse. Experiments with eight memory systems across two backbones show that current systems often fail to use observed evidence or turn feedback into reliable follow-up behavior, even when evidence is stored or feedback is incorporated locally. StreamMemBench is publicly available at https://github.com/landian60/StreamMemBench.

Emulators provide a cost-effective alternative to regional climate models (RCMs) by capturing their dynamical downscaling function. They link large-scale predictors simulated by global climate models (GCMs) to RCM-simulated high-resolution fields of the target variable, here precipitation. Machine learning methods, typically deep learning, are cheaper than running RCMs in computation time and energy. Among them, generative models are appealing because they can simulate ensembles of local high-resolution fields consistent with the predictors. This ensemble, which we call the uncertainty envelope, remains to be properly assessed for added value. Here, we make three contributions. First, we introduce ParamDiffusion, a new two-stage diffusion-based framework, and compare it with a state-of-the-art diffusion approach. Second, we expand standard validation through a comprehensive framework aligned with climate-science needs, examining specific precipitation events, including extremes. Third, within this framework, we assess the added value of diffusion approaches relative to deterministic methods. We intercompare four deep-learning models: a deterministic model designed to capture the precipitation tail; a parametric probabilistic model based on it; a recently proposed diffusion approach; and ParamDiffusion, which couples the parametric model with a diffusion model. Our results show that diffusion-based approaches reproduce climatological precipitation statistics with high skill, including distributional tails and spatially compounded extremes, while generating spatially detailed fields. However, none of the assessed models consistently accounts for the most extreme RCM-simulated events within its uncertainty envelope. Diffusion models are therefore promising for probabilistic RCM emulation, but progress is still required before they can reliably represent high-impact precipitation extremes.

Cycle-level DRAM simulators provide accurate and flexible models for DRAM and memory controller operations and enable research on current and future memory systems. Therefore, they are critical for improving the performance, efficiency, and robustness of DRAM-based memory systems. Ramulator 2.0 (successor of Ramulator) is a highly modular and extensible cycle-accurate DRAM simulator that enables rapid exploration of new ideas in DRAM-based memory systems. A MICRO 2024 best paper runner-up publication, A Mess of Memory System Benchmarking, Simulation and Application Profiling, which we refer to as "the Mess paper," with all three artifact badges awarded (including "Reproducible"), proposes a new benchmark to evaluate real and simulated memory system performance. While doing so, it makes strong negative claims about Ramulator 2.0 and shows unexpected results. In this talk and the associated extended abstract, we demonstrate that these results and claims in the Mess paper are incorrect and are due to configuration and simulator usage errors made in the Mess paper. We describe four best practices to aid users and developers of simulation tools to avoid such issues in the future. We emphasize the importance of contacting simulator authors and developers when unexpected results are observed (especially and importantly before publishing such results), to ensure these simulators are used with correct configurations and as intended. Our investigation also aims to stimulate discussion on artifact evaluation practices and on mechanisms for correcting results and artifacts after publication. To aid future works and reproduction of all our results, we open source all our code and scripts at https://github.com/CMU-SAFARI/Cleaning-up-the-Mess. We refer the reader to our full ISPASS 2026 paper and its artifact for the complete analysis, detailed methodology, and extended results.

Constitutive artificial neural networks (CANNs) provide interpretable material model discovery, but have so far been used in stress-supervised settings based on apparent stress-strain data from homogeneous tests. Because each test samples only a narrow loading path and provides homogenized rather than local stress information, robust discovery typically requires multiple loading modes to constrain the multidimensional response. This is challenging for soft biological tissues, where repeated testing, damage, and sample variability limit reliable information from a single specimen. Here, we combine CANNs with the stress-unsupervised full-field discovery framework EUCLID to identify sparse hyperelastic laws directly from displacement fields and reaction forces in one heterogeneity-inducing loading case. CANN-EUCLID minimizes equilibrium imbalance with sparsity-promoting regularization selecting compact active terms, without local stress measurements or a prescribed law. We evaluate the approach on isotropic and anisotropic benchmarks with prescribed ground-truth laws. When the ground truth is representable by the chosen CANN basis, our method recovers the correct terms with near-exact accuracy, including exponential terms with embedded parameters. When it is not contained in the basis, the method retains shared terms and approximates missing contributions using available basis functions. Generalization depends strongly on sampled deformation states: exponential strain-stiffening terms can be recovered accurately when sufficiently probed, but can produce large extrapolation errors when the stiffening regime lies outside the sampled domain. Forward FE validation simulations show that the discovered behavior accurately replicates the ground truth. These results establish stress-unsupervised CANN discovery as a promising framework for interpretable full-field constitutive model identification.

Sociable weaver nests function as complex ecological structures offering thermoregulatory microhabitats and sustaining diverse species; however, datasets used in prior studies lack fine-grained 3D structural detail. Producing usable and accurate 3D weaver nest data is challenging due to their irregular geometry and integration with complex host vegetation. We bridge this gap with an open-access, 1.4 TB multimodal drone dataset of 104 nest-bearing trees, comprising 27,945 RGB images, 111,780 multispectral images, approximately 781 million 3D points, and expert-annotated semantic segmentation labels. We benchmark semantic segmentation using KPConv, RandLA-Net, and Point Transformer V3, with PT-v3 achieving an mIoU of 86.35% on the test set. While the results demonstrate strong performance for transformer-based and point-wise methods, they also highlight architecture-dependent challenges, particularly for convolution-based approaches such as KPConv. By uniquely combining spectral, spatial, and structural information, the presented dataset advances 3D reconstruction, segmentation, and classification algorithms, enabling ecological applications from nest volume estimation to species conservation, and serves as a demanding benchmark that exposes architecture-dependent performance under extreme class imbalance.

Robotics manipulation research increasingly focuses on two-finger parallel grippers for their effectiveness, affordability, and ease of teleoperation. Grippers are nonetheless limited by their form factor, often requiring bimanual setups even for simple reorientation tasks. Anthropomorphic hands are a more natural platform for dexterous robot learning -- closer to the human hand, and capable of learning from human video -- yet they remain hard to use in learning research: even where open and accessible hand hardware exists, the software for control, simulation, teleoperation, and retargeting is scattered in one-off code bases, and largely disconnected from the robot-learning ecosystem. In this work, we introduce the \orca~learning stack, an open-source research stack for dexterity as a first-class robot learning domain. Our \orca~stack unifies low-level control, simulation, teleoperation from a range of consumer platforms, and hand retargeting, behind a single interface, and integrates natively with popular robot-learning frameworks such as \lerobot, so dexterous hand researchers can leverage the same data, training, and evaluation pipelines used for non-dexterous robot learning. We demonstrate a complete end-to-end workflow, collecting expert demonstrations of an in-hand reorientation task by teleoperation with a consumer-grade VR headset, training an autonomous policy with \lerobot, and evaluating the learned policy in a fully reproducible and observable setup. We open-source the entire stack as a shared, reproducible foundation for dexterous-manipulation research.

Non-Euclidean optimisation methods with matrix-valued updates, such as Muon and Scion, have recently shown strong empirical performance for training Transformer models, yet their theoretical advantages over Euclidean methods remain poorly understood. We address this gap in the heavy-tailed non-convex regime, where stochastic gradients have bounded $p$-th central moments, $p \in (1,2]$. We show that certain non-Euclidean methods achieve optimal sample complexity under stronger stationarity measures, while Euclidean methods incur additional dimension-dependent costs. As a consequence, for $m \times n$ matrices, Muon finds an $\varepsilon$-stationary point in nuclear norm within $\mathcal{O}\left(\min\{m, n\} \frac{Δ_1 L}{\varepsilon^2} \left(\frac σ\varepsilon \right)^{\frac p {p-1}}\right)$ samples, absorbing heavy-tailed noise without extra dimension dependence, unlike Euclidean methods. We further prove this sample complexity, including its dimension dependence, is optimal for all first-order methods under nuclear-norm stationarity. Experiments on large language models support our theory. Surprisingly, our results suggest that other Schatten geometries beyond the spectral geometry of Muon can perform competitively in certain settings.

Modern image classifiers widely adopt global average pooling (GAP) followed by a linear classification head. This linearity ensures that the image-level logits equal the average of logits obtained by applying the classification head pointwise to the feature grid prior to GAP. Consequently, standard classifiers may inherently retain spatial class evidence that remains recoverable even when the image-level prediction is incorrect. This structure naturally suggests a multiple-instance learning (MIL) interpretation, where an image is viewed as a bag of spatial instances. Within this formulation, we demonstrate that standard classifiers trained with a single label per image can still learn the intended classification task in multi-object scenes. We further exploit this property to decompose image-level logits into a prediction grid, providing a post-hoc diagnostic to extract spatial class evidence that GAP otherwise obscures. Our systematic evaluation reveals that off-the-shelf models consistently recover the ground-truth class within foreground regions. The MIL interpretation further suggests that common classifier failures reflect known limitations of mean aggregation.

Blockchain interoperability enables independent blockchain systems to communicate and exchange assets across heterogeneous networks. However, the lack of comprehensive security mechanisms remains a critical weakness -- one that attackers have already exploited to cause hundreds of millions of dollars in asset losses. This paper presents a systematic identification and classification of security threats facing interoperable blockchain systems, along with corresponding countermeasures for each. We organize threats into five categories: (1) core blockchain attacks, (2) network attacks, (3) interoperability-specific attacks, (4) social engineering, and (5) code vulnerabilities, with particular attention to smart contract weaknesses. For each identified threat, we analyze its attack surface and propose effective defensive strategies. The resulting taxonomy provides a structured foundation for designing and evaluating secure blockchain interoperability solutions.

Robots under autonomous operation may require decisions based on evidence that is no longer visible. We study \emph{delayed-evidence} tasks, where an early cue disappears before a later decision point, so visually similar observations can require different actions. In these settings, the current observation is not a sufficient state for control. We introduce TRAjectory-routed Causal Evidence (TRACE), a memory framework for visuomotor imitation policies. TRACE stores task-relevant visual and robot-state evidence, such as object identity, target choice, or route-dependent state, in a fixed-size latent memory that remains bounded over long episodes. Instead of indexing memory by raw time or manually provided task labels, TRACE uses \emph{path signatures}: compact, order-sensitive features of the executed robot-state trajectory. These signatures do not store the visual cue itself; rather, they provide trajectory-conditioned keys for writing and retrieving the evidence stored when the cue was visible. When the robot later reaches an ambiguous observation, the policy conditions on TRACE memory to recover the missing context and choose the correct branch. TRACE attaches through lightweight adapters to policies, without changing the policy backbone, action head, or imitation objective. Across real-world long-horizon manipulation tasks with visually ambiguous branch points, TRACE improves branch selection and task success over alternative baselines, including short-history and recurrent memory. Project page: https://jeong-zju.github.io/trace

Safe reinforcement learning (RL) aims to learn policies that optimize rewards while satisfying constraints. Predominant approaches rely on soft-constrained policy optimization, which has achieved empirical success but does not provide formal safety guarantees for the learned policy. In contrast, methods with strict guarantees typically rely on explicit certificate functions, whose construction requires the direct synthesis and verification of control-invariant sets, a process that scales poorly with state dimension and often yields overly conservative behavior. In this paper, we present the Provably Safe, yet Scalable RL (PS2-RL) framework, a novel two-phase architecture for learning provably safe policies in a scalable manner, designed to overcome the key bottlenecks of prior methods. Rather than explicitly computing invariant sets, PS2-RL leverages a learned backup policy to forward-integrate the system dynamics, generating an implicit control-invariant set online. In the first phase, the backup policy is trained with our proposed safe-arrival value function, which characterizes the optimal backup policy for invariant-set construction. In the second phase, an RL policy is trained end-to-end through a differentiable projection layer that strictly enforces the safety guarantees induced by the learned backup policy. By maximizing the volume of the implicit control-invariant set in the first phase, the resulting PS2 policy from the second phase is performant and scalable, while maintaining provable safety. Crucially, PS2-RL imposes no restrictions on the underlying RL algorithm and can be plugged into any existing training pipeline. We establish theoretical guarantees for the proposed framework and evaluate it on robotic control tasks with state dimensions up to 10, a regime in which prior provably safe RL methods struggle or become impractical.

Principal Component Analysis (PCA) preserves variance, not the information needed to detect rare catastrophic events. This paper proves the existence of a {\it Risk Shadow}: PCA can retain over 99.9999 percent of total variance while completely erasing all signal about rare, high-impact failures. When this happens, even the best possible classifier operating on the PCA representation reduces to a constant predictor. The root cause is a fundamental mismatch between variance maximization and tail risk awareness. To break the shadow, we introduce Expectile PCA (ExPCA) and Tail-Preserving PCA (TP-PCA), two methods that reweight the data covariance toward high-impact events. We prove theoretically that ExPCA strictly outperforms PCA in retaining rare-event information, and we validate our claims on synthetic data and a real-world credit card fraud detection benchmark. Our results call for a fundamental rethinking of variance-based dimensionality reduction in high-stakes decisions.

Large language models encode rich information in their hidden states. This work asks whether code correctness is legible in the hidden states of Qwen3-4B-Instruct-2507, before it generates and as it repairs a failed attempt, studied on 444 LiveCodeBench tasks. It reports two findings connected by a single confound-control tool: residualization. First, the correctness of the model's first-attempt code is linearly decodable from the prompt-final hidden state, with a leakage-free held-out AUC of 0.931 +/- 0.008 across 50 outer splits. After the linear effect of prompt length is removed from each hidden state dimension, the probe still reaches 0.911 +/- 0.010, well above a prompt-length baseline of 0.754 +/- 0.014. Second, on 236 cleaned cases where the model attempts to repair a failed first attempt, the hidden state shift from the failing attempt to its repair carries a statistically detectable contrastive direction, significant on both a magnitude and a split-half test against label-shuffled nulls. This direction does not survive a conditional residualization against repair-context covariates that differ between successful and failed repairs, marking it as a correlate of repair success driven by the repair context rather than an isolated repair-comprehension feature. The probe layer is selected by nested cross-validation, and the same residualization approach that upholds the pre-generation correctness result overturns the repair-direction interpretation. The contribution is as much methodological as empirical: a diagnostic honest enough to report a negative result alongside a positive one.

Real-time, full-duplex speech interaction is a key feature of next-generation spoken chatbots, allowing the model to listen and speak at the same time and to handle natural phenomena such as overlap, hesitation, and barge-in. Existing speech language models (SpeechLMs) such as LLaMA-Omni and GLM-4-Voice are still turn-based and rely on an external Voice Activity Detection (VAD) module to mark the end of the user's turn, which fundamentally limits their interactive ability. In this paper, we introduce BayLing-Duplex, a native full-duplex SpeechLM where a single autoregressive LLM decides when to listen, when to speak, and when to stop, with no auxiliary turn-taking module. The design adds only a few special tokens to the standard vocabulary, so it transfers across LLMs and reuses existing training and serving stacks with no architectural adaptation. Starting from the public GLM-4-Voice checkpoint and using only 400K full-duplex samples for fine-tuning followed by a lightweight DPO stage, BayLing-Duplex reaches 92% turn-taking success and 100% interruption success on InstructS2S-Eval, while improving the speech-response score from 2.17 to 3.39 over Moshi. BayLing-Duplex also matches or surpasses its turn-based counterpart on Llama Questions, Web Questions, and Alpaca-Eval, showing that simultaneous listen-and-speak modeling does not sacrifice response quality.

The removal of learned data from Machine Learning models through Machine Unlearning (MU) has been widely studied; however, there has yet to be an agreed-upon scheme for auditing MU. Existing work has shown that a dishonest model owner can falsify evidence to avoid executing MU, while curious auditors (and adversaries) can infer the privacy-sensitive properties of the model and its training data even with limited access. Yet auditing of MU under mutual distrust between the model owner and the auditor remains unexplored. We provide an information-theoretic proof for this scenario: for convex ML models, a generic audit scheme that relies solely on querying the model for \textit{behavioral} signals cannot identify insufficiently unlearned models without revealing membership information of the retained set. Therefore, auditing MU under the assumption of a dishonest model owner and an honest-but-curious auditor faces an inherent privacy-audit tradeoff. Our empirical results on convex models strongly supports this result, while further experiments demonstrate that this privacy-audit tension persists in non-convex models. Our results call for a more careful consideration of the privacy-audit tension under a realistic auditor threat model, and serve as a foundation for more scrutiny of designs of privacy-preserving audit schemes for the MU pipeline. We also release our code implementation at https://github.com/LiouTang/Behavioral-Unlearn-Audit.

LLM-based guardrails have emerged as a highly effective defense against prompt injection and jailbreak attacks in autonomous agents. However, we reveal that the very reasoning and task-following capabilities enabling this protection introduce a novel vulnerability: attackers can inject crafted data to trap the guardrail in extended reasoning loops, effectuating a systematic denial-of-service (DoS) attack. To systematically expose this threat, we design a beam-search optimization framework that crafts natural-language payloads to maximize guardrail reasoning length, utilizing an LLM proposer guided by a strategy bank. Based on the observation of guardrail's schema-following nature, we also provide another attack framework driven by mechanism-aware structural mutations with less computational load. The attack efficacy is systematically evaluated in two parts. First, in standalone evaluations, the attack generalizes across diverse guardrail architectures, safety templates, and agent benchmarks. Payloads optimized on a single open-source surrogate successfully transfer to eight leading model backbones (e.g., Claude, GPT, Gemini, DeepSeek, and Qwen), achieving a 13--63$\times$ token amplification. Second, in end-to-end real-world agent deployments (web, desktop, code, and multi-agent systems), the attack reveals up to a 148$\times$ latency amplification. We show that a single poisoned document can saturate shared guardrail infrastructures, effectively starving co-located agents and paralyzing the entire system. By uncovering this availability flaw, our work underscores the urgent need to develop cost-bounded, reasoning-robust guardrails.

AI evaluations are widely used for testing and understanding progress. However, the diverse evaluators bring with them inconsistencies that challenge analysis and comparison. First, results are saved in incompatible formats, scattered across leaderboards, papers, blog posts, evaluation harness logs, and custom repositories. Second, results are created by different evaluation frameworks, which produce divergent scores for nominally identical evaluations and record metadata inconsistently, hindering comparison, cross-community evaluation science, cost reduction, and reuse. We introduce Every Eval Ever, the first shared schema and community-crowdsourced repository for AI evaluation results. The schema standardizes how evaluations are represented in a unified, single JSON document. It is source-agnostic by design, ingesting results from evaluation harnesses and papers alike, and optionally stores per-instance outputs for fine-grained analysis. We contribute: (i) a community-governed metadata schema with a companion instance-level schema, the first standardization effort of its kind; (ii) automatic converters from popular formats, evaluation harnesses, and leaderboards to the unified schema; and (iii) a crowdsourced community database hosted on Hugging Face, currently spanning to date 22,235 models, 2,273 unique benchmarks, and 31 evaluation formats.

Internet of Medical Things (IoMT) devices operate under strict resource constraints while handling highly sensitive health data, making security and privacy critical concerns. Federated learning (FL) further complicates this landscape, as model updates exchanged during training may unintentionally expose private medical information. Emerging quantum computing capabilities threaten the long-term viability of conventional lightweight cryptographic mechanisms, motivating the integration of Post-Quantum Cryptography (PQC) into IoMT systems. This article discusses key enabling technologies for quantum-resilient IoMT, including post-quantum key establishment, lightweight encryption, and edge-native orchestration. We propose a scalable Kubernetes-based framework that integrates PQC into FL-enabled IoMT environments and validate it on a Raspberry Pi testbed. Results demonstrate that distributed cryptographic processing significantly reduces latency compared to sequential designs while maintaining feasible resource overhead. The primary contribution of this work lies in the design and validation of a secure orchestration and communication framework for FL-enabled IoMT systems. We conclude by outlining future directions toward energy-aware architectures, intelligent security optimization, and resilient next-generation Intelligent Internet of Medical Things (IIoMT) ecosystems.

The recent successes of neural networks producing human-like language have caused significant stir in cognitive science, with many researchers arguing that classical puzzles about human cognition and challenges to artificial intelligence are being solved by neural networks. A notable case is the argument from systematicity due to Jerry Fodor and Zenon Pylyshyn, argues that humans display systematic biconditional dependencies. For example, someone can understand the sentence "John saw Mary" just in case that they understand the sentence "Mary saw John." Symbolic systems explain this systematicity of language and thought, while neural networks offer no immediate explanation. Several recent articles argue that this challenge has now been met by neural networks. In particular, Brenden Lake and Marco Baroni argue that their meta-learning for compositionality protocol matches and perhaps explains human systematicity. We demonstrate that these conclusions are premature. Among other results, we found that their model struggles to learn rules that are even slightly out of distribution compared to their training data. Furthermore, the model behaves unsystematically even on many within-distribution problems. We conclude that Fodor and Pylyshyn's challenge to neural networks remains unmet.

Macrocyclic peptides are promising therapeutic candidates for intracellular targets, but their design requires simultaneous control over non-natural monomer chemistry, ring topology, membrane permeability, and target binding. Existing SMILES- or HELM-string generative models either operate in long atom-level sequence spaces or treat monomers as symbolic tokens with limited chemical grounding. We introduce PepALD, an Autoregressive Latent Diffusion (ALD) foundation model for \textit{de novo} macrocyclic peptide generation. The model represents HELM monomers with structured chemical embeddings, generates each residue through context-conditioned diffusion in chemically informed latent space, predicts R-group-aware ring closures during autoregressive generation, and aligns the denoiser to affinity rewards using winner-protected diffusion-adapted preference optimization. In silico experiments demonstrate PepALD's generation quality and reward-optimization performance against representative peptide generation baselines.

Fine-tuning vision-language models to emit dense coordinate lists improves visual grounding but also changes how models serialize, repeat, and terminate structured outputs. We study this behavior as a generation and control surface. In Gemma 4 12B, high-capacity q/k/v/o LoRA raises class-aware F1@0.3 from 0.007 to 0.448 while inducing repeated-tail pressure (duplicate rate 0.080, max repeat 23). A q/v rank sweep keeps max repeat at 21-22 across ranks 4-64, showing capacity persistence. The target signal is separable: object-level repeat-stop removes exact repeated records (duplicate rate 0.000, max repeat 1) while preserving F1 (0.494 to 0.490) and stricter F1@0.5 (0.381 to 0.385). Structure-axis probes localize the effect to bbox-coordinate object lists; dense non-bbox and spatial/count JSON remain repeat-clean, including under high-capacity adapters. Qwen3-VL-8B reproduces a clean controlled endpoint (F1@0.3 0.318, duplicate rate 0.000), and COCO 2017 reproduces acquisition plus duplicate pressure. Dense coordinate-list adaptation therefore creates a structure-bound, cross-family interference surface that can be measured and controlled.

Understanding how a prediction model will perform in a new environment before deployment is essential to preventing harm when algorithms inform decision-making. Two common sources of model performance degradation are (i) covariate shift, where the target covariate distribution differs from the source, and (ii) selective labels, where the observability of outcomes depends on historical decisions. We study pre-deployment model evaluation under the joint presence of covariate shift and labeling of outcomes selectively based on observed features. In particular, we present a double machine learning procedure for estimating the target risk of an arbitrary black-box prediction model under a general loss function. We show identification of this estimand under standard assumptions and derive a bias-corrected estimator based on the influence function of the target risk. Finally, we evaluate our estimator through experiments using the eICU electronic health records database, showing that it tracks the true target risk more accurately than methods that address either selective labels or covariate shift alone, as well as baselines that combine standard plug-in approaches.

Large Language Models (LLMs) are undergoing a fundamental transformation from conversational generators into integrated AI systems capable of reasoning, action, memory, and self-improvement. We conceptualize this transition as a shift from Chatbot to Digital Colleague: from conversational answers to persistent work. We organize this transition along two tightly coupled dimensions. First, at the cognitive core level, LLMs are advancing from Chatbot-era "fast thinking" systems driven by next-token prediction toward Thinking LLMs that leverage inference-time computation, Chain-of-Thought reasoning, reflection, process supervision, and reinforcement learning to support more deliberate and reliable cognition. Second, at the tool-augmented task execution level, LLMs are progressing from tool-calling Agents that invoke external resources in an ad hoc manner toward OpenClaw-style workstation systems (OpenClaw) equipped with persistent Workspaces, skills, verification loops, and governance. The "Workspace + Skill" paradigm makes episodic tool use colleague-like via state persistence, reusable procedures, task closure, and experience reuse. We examine data construction shifts from instruction-response pairs to State-Action-Observation trajectories and evaluation from static benchmarks to sandboxed, auditable, self-evolving AI ecosystems.

Predicting the Kohn-Sham Hamiltonian with machine learning can accelerate density functional theory while retaining access to molecular orbitals, energy levels, and electronic-structure observables that energy-only surrogates cannot resolve. Yet element-wise agreement with the converged Hamiltonian, an implicit fixed point of the self-consistent field iteration, does not determine the occupied subspace that governs orbital energies and densities. Here we present HamEvo, a neural operator that learns the single-step self-consistent update and returns the converged Hamiltonian as its fixed point. HamEvo is pre-trained on intermediate self-consistent trajectories and calibrated at equilibrium with density-matrix supervision. Across benchmarks from MD17 to drug-like QMugs, HamEvo lowers Hamiltonian errors by 35-49% over direct-regression and deep-equilibrium baselines, and predicts QMugs HOMO and LUMO energies with mean absolute errors of 0.036 and 0.053 eV, near the 1 kcal/mol chemical-accuracy scale. Few-shot fine-tuning with only 20 reference conformations extends HamEvo to molecules of up to 122 atoms, well beyond the size range covered by pre-training. With thermal molecular-dynamics sampling, HamEvo captures temperature-dependent HOMO-LUMO gap renormalization beyond the harmonic approximation. Inference is up to 242 times faster than conventional DFT.

We present a recipe-controlled decoder audit (RCDA) for structural transductive knowledge-graph completion (KGC). The audit asks a simple reporting question: before attributing gains to an encoder or training recipe, what changes when the decoder is swapped under the same recipe? Using ComplEx and DistMult as the primary controlled pair, with targeted RotatE/TransE spot-checks, we evaluate seven benchmarks. On five standard KGs, ComplEx-vs-DistMult differences are modest but consistent under our recipe (+0.005 to +0.012 MRR), whereas CompGCN-style encoder effects vary more by dataset. On small KGs, decoder effects become the main diagnostic: Kinship shows a stable ComplEx advantage of +0.143 MRR (6 seeds), while UMLS favours ComplEx by +0.022 MRR in a clean 6-seed server rerun but reverses in an earlier provenance variant. We therefore treat small-KG decoder choice as recipe- and provenance-sensitive rather than as a fixed dataset winner. We further show that decoder choice interacts with encoder depth on WN18RR, and that under our recipe L=0 ComplEx on YAGO3-10 reaches 0.6971 +/- 0.0048 MRR at d=128. The result is a compact audit protocol: report matched decoder rows, log small-KG provenance, and sweep decoder x depth before making encoder-level claims.

Recent finite-time analyses of nonlinear two-time-scale stochastic approximation show that under contractive assumptions the slow iterate $Y_k$ with stepsizes $β_k=Θ(k^{-1})$ and $α_k=Θ(k^{-a})$, $a\in(1/2,1)$, generally satisfies a mean-square rate of order $k^{-a}$; decoupled $k^{-1}$ rates require strong local linearity. We identify a sharp regularity-dependent boundary. In a rate-determining normal form where the slow drift contains a locally linear leakage and a nonlinear remainder of order $1+ρ$ ($ρ\in[0,1]$), the uncorrected recursion satisfies \[ \mathbb{E}\|Y_k\|^2 \le C\bigl(k^{-1}+k^{-a(1+ρ)}\bigr), \] and a matching scalar Gaussian lower bound shows that the slower term is unavoidable without modifying the update. Thus the decoupled $k^{-1}$ rate is guaranteed for the uncorrected recursion exactly when $a(1+ρ)\ge 1$. This lower bound concerns only the naive update; it is not an information-theoretic obstruction. We demonstrate this by equipping the normal-form recursion with an auxiliary online bias estimator \[ M_{k+1}=M_k+γ_k(R(X_k)-M_k),\qquad β_k\llγ_k\llα_k, \] and subtracting $M_k$ from the slow update. Under the same stability, moment, and remainder assumptions, the corrected recursion achieves $\mathbb{E}\|\widetilde Y_k\|^2=O(k^{-1})$ for every $ρ\in[0,1]$, including regimes where the uncorrected update provably suffers the slower rate. Finally, we prove localized transfer theorems that extend the phase-transition mechanism to general nonlinear TTSA in fast-manifold coordinates. The proofs are non-asymptotic and rely on two Abel-transform cancellations: one for the locally linear fast-error leakage, and one for the tracked nonlinear bias.

A growing line of work equips large language model (LLM) agents with graph neural networks (GNNs) as callable tools, assuming the agent exercises judgment over when and how much to rely on such a tool. We test this directly. We expose a frozen GNN to a ReAct-style LLM agent as an explicit tool and measure, on node classification over a text-attributed graph (ogbn-arxiv, replicated on WikiCS), whether the agent uses the tool or merely obeys it. We find the agent does not exercise judgment: its predictions agree with the raw GNN's 97.6-99.2% of the time (5 seeds), collapsing into a GNN parrot that adopts the tool's output wholesale and bypasses its own reasoning. Sweeping backbone capability (Qwen2.5 0.5B-7B), the deference is not a weak-model artifact: among models able to invoke the tool, agreement rises with capability (0.60 to 0.98 from 1.5B to 7B). Crucially, the cost of deference does not shrink as capability grows and grows where alternatives emerge: a per-node oracle over the available actions beats the parrot by 0.09-0.18 at 3B and 0.12-0.22 at 7B, roughly doubling at high homophily, because the parrot is pinned to the frozen GNN while the agent's alternatives improve; at 7B a simple neighbour-label tool overtakes the GNN at high homophily (0.81 vs 0.71) yet the agent still defers. A simple selective-invocation gate recovers about half of that high-homophily gap (0.71 to 0.83) but yields no net global gain, and held-out estimates bound the best achievable gate over standard test-time features to at most a third of the oracle headroom: reliable selective invocation looks limited by available information, not merely router design. Our results are a cautionary measurement: evaluations of agent+tool systems cannot assume the agent adds judgment on top of the tool, and selective invocation must be designed in rather than expected to emerge from scale.

Large language model (LLM) reasoning is ephemeral: chains of thought vanish with the context window, pruned search branches leave no record, and memory buffers cannot be diffed, merged, or audited. Every other complex software process (code, infrastructure, data, experiments) is version-controlled; reasoning is not. We introduce GitOfThoughts, which stores an agent's reasoning tree as a git repository: every scored thought is a commit, scores are notes, outcomes are tags, and retrieval is "git log" over the agent's own history. This makes reasoning replayable, auditable, and mergeable across agents at near-zero engineering cost. We then ask the harder question: does memory, in any substrate, actually improve accuracy? Across five substrates (none, markdown, vector, graph, git), two benchmarks, two model scales, and pre-registered replications, the answer for novel problems is no. No memory format reliably helps, and a promising early result collapsed under its own pre-registered replication. Memory pays only above what we call the copyability threshold: when the retrieved case is a near-duplicate of the current problem (similarity >~ 0.8), accuracy jumps sharply; below it, nothing. The gain is answer retrieval, not method transfer: a 4.5x larger model doubles the near-duplicate payoff yet still cannot extract a transferable method from a worked example. The only general lever we find is test-time sampling. The case for git-as-substrate is therefore auditability, provenance, and mergeability at accuracy parity. We document a retracted result and a refuted hypothesis to model the evaluation standard we hold ourselves to.

This paper investigates the fragility of post-hoc explanation methods in audio deepfake detection. While previous work on explanation manipulation focused on images using standard $L_p$ metrics, we introduce a psychoacoustic framework that optimizes inaudible perturbations to decouple model attributions from final classifications. We evaluate this vulnerability across state-of-the-art architectures under strict prediction-preserving constraints. By evaluating the manipulation cost through domain-specific perceptual audio quality metrics alongside explanation alignment criteria, our framework demonstrates that an adversary can systematically distort automated explanation heatmaps while preserving the predicted deepfake label. Full code available at: https://github.com/cncPomper/Audio-XAI

Neurosymbolic (NeSy) models integrate neural networks and symbolic reasoning for robust and interpretable AI. State-of-the-art NeSy models require that the symbolic component is expressed in a differentiable way, often complicating the use of approximate inference. We propose EM-NeSy which casts probabilistic NeSy learning as an instance of the Expectation-Maximization (EM) algorithm. In the expectation step, we compute the posterior over the neurally predicted symbols conditioned on the label via probabilistic inference. In the maximization step, we update the neural parameters based on this posterior using gradient descent only through the neural component. This formulation unlocks the full potential of the EM algorithm for NeSy learning. It allows NeSy to extend naturally to approximate reasoning without any additional modifications or differentiability requirements of the symbolic component. Furthermore, it recovers the standard end-to-end gradient-based NeSy setting under exact inference. Our experimental results demonstrate the scalability and computational efficiency of EM-NeSy.

Transformer-based clinical language models are increasingly integrated into high-stakes clinical decision support pipelines, yet the computational mechanisms through which demographic associations encoded in medical documentation propagate into model probability distributions remain empirically underspecified. We present a systematic computational audit of representational bias in ClinicalBERT (Alsentzer et al., 2019), a BERT-based model pretrained on MIMIC-III discharge summaries, employing two complementary probing methodologies: Log Probability Bias Analysis (LPBA), which quantifies demographic descriptor-induced shifts in masked token probability distributions across behavioral and evaluative semantic categories, and Masked Language Model-based analysis (MLM), which probes internal representational structure for demographic agency attribution encoding across 98 real clinical sentence templates and eight intersectional race-gender combinations. Corpus frequency analysis operationalizes the distinction between statistical disparity and bias amplification by benchmarking model outputs against empirical term frequencies in the MIMIC-III training corpus. Of 32 statistically significant findings, 65.6% contradict observed corpus distributions, rising to 80% for Black patients and 87.5% for agency attribution under MLM probing, providing direct empirical evidence that representational bias in ClinicalBERT operates predominantly through model-internal amplification rather than training data inheritance. Keywords: natural language processing, clinical documentation, algorithmic auditing, representational bias, health equity 1

Modern Automatic Speech Recognition (ASR) systems have made remarkable progress on standard benchmarks, yet performance gaps have emerged under real-world distribution shifts, caused by recording conditions, accents, speech impairments, and noise. Existing datasets and benchmarks typically isolate these factors, which overlooks their co-occurrence in real-world applications. In this paper, we argue that model robustness can be treated as a dynamic capability that continually develops, and we introduce MoDiCoL, a Modular Diagnostic Continual Learning dataset designed for controlled analysis of linguistic content, speaker characteristics, and acoustic environments. Furthermore, we propose a real-world-inspired continual learning curriculum to simulate incremental updates and study how robustness is acquired, transferred, and forgotten. We evaluate three continual learning strategies and provide detailed insights into robustness under evolving conditions.

Existing multi-agent software development systems have proposed many forms of agent collaboration, including role-based collaboration and automated code review. However, many systems assume a common runtime, a central conversation server, or the same API family. Under these assumptions, LLM agents from different vendors cannot easily exchange messages directly from their own execution environments while dividing development and review work on a shared codebase. This paper presents tap, a file-based collaboration protocol that allows Claude (Anthropic) and Codex (OpenAI) to collaborate on one codebase without shared memory or an identical runtime. The core of tap is a file-first design that preserves markdown files with metadata as original messages, combines a file inspection path (file communication, Tier 1) with real-time notification paths for Claude and Codex (real-time communication, Tier 2), and isolates work through separate git worktrees. Even if real-time notification fails or a receiver restarts, the message file remains available and the same content can be inspected again. In a 27-day, 37-generation self-applied operation where tap was used to develop and review itself, we collected 209 tap-related pull requests and 717 operational artifacts. An analysis of 375 review artifacts showed that the share of reviews recording at least one defect or requested change was 69.8% for heterogeneous model pairs and 53.1% for homogeneous model pairs. These results show that tap, which combines file-based message preservation with real-time notification, operates in a real production repository, and that combining heterogeneous models and execution environments can broaden review perspectives. tap is distributed as the open-source npm package @hua-labs/tap (v0.5.2).

End-to-end (E2E) autonomous-driving planners trained by imitation are prone to statistical shortcuts: they associate scene elements that merely co-occur with expert actions (a roadside object, a building facade) with driving decisions, rather than the variables that causally determine them. Such causal confusion silently compromises reliability in long-tail scenarios, and it is difficult to detect, because prevailing open-loop metrics (L2 displacement and collision rate) are dominated by ego status and do not indicate whether a planner depends on spurious cues. Existing remedies based on causal-intervention training require retraining large models and cannot audit a planner that is already deployed. We present CADET, a training-free framework that audits, benchmarks, and repairs spurious reliance in pretrained E2E planners without any parameter update.

How do people cope when disaster strikes and can we detect it at scale, in real time, from what they write? This study addresses that question using over one million Turkish-language tweets posted in the aftermath of the February 6, 2023 earthquake in Turkiye, which unfolded in a deeply polarized political context just months before a national election. Drawing on Lazarus and Folkman's (1984) coping theory, we develop a multi-label BERTurk classifier to detect three coping styles (problem-focused, emotion-focused, and meaning-making) across four theoretically motivated crisis phases. BERTurk achieves a macro F1 of 0.693, substantially outperforming a zero-shot mDeBERTa baseline (macro F1 = 0.324). Applied to the full corpus, the classifier reveals a clear temporal trajectory: problem-focused coping dominates the urgency phase and declines sharply, emotion-focused coping rises and stabilizes, and meaning-making increases monotonically. Anger correlates most strongly with meaning-making (Spearman r = 0.387), suggesting it functions as a mobilizing force toward blame attribution rather than practical action. These findings demonstrate that coping theory can be reliably operationalized in real-world digital crisis data and that doing so can help humanitarian organizations tailor their responses to where a population actually is.

We introduce COMET (Causal Object-centric Model for Efficient Tree search), a model-based reinforcement learning algorithm that performs Monte Carlo Tree Search in a slot-structured latent space. COMET pairs a frozen unsupervised object-centric encoder with a transformer-based world model, in which actions are bound to objects through a novel action-slot fusion mechanism that is used in slot transition prediction. Policy and value heads use object-causal attention, modulating token interactions by learned per-slot relevance scores so that decision-making concentrates on task-relevant entities. COMET adds an explicit object-level inductive bias to MuZero-style latent planning. Across eight visually and dynamically diverse tasks from the Object-Centric Visual RL benchmark, ManiSkill, Robosuite, and VizDoom, COMET achieves a higher mean normalized score during the early stages of training compared to object-centric and monolithic baselines.

Federated learning (FL) often struggles with generalization due to heterogeneous client data. Local models are prone to overfitting their local data distributions, and even transferable features can be distorted during aggregation. To address these challenges, we propose FedCONST, an approach that adaptively modulates update magnitudes based on the parameter strength of the global model. This prevents over-emphasizing well-learned parameters while reinforcing underdeveloped ones. Specifically, FedCONST employs linear convex constraints to ensure training stability and preserve locally learned generalization capabilities during aggregation. A Gradient Signal to Noise Ratio (GSNR) analysis further validates the effectiveness of FedCONST in enhancing feature transferability and robustness. As a result, FedCONST effectively aligns local and global objectives, mitigating overfitting and promoting stronger generalization across diverse FL environments, achieving state-of-the-art performance.

Safe reinforcement learning (Safe RL) aims to maximize expected return while satisfying safety constraints, typically modeled as Constrained Markov Decision Processes (CMDPs). While primal-dual methods scale well to deep RL, they often suffer from delayed constraint correction, leading to oscillatory behavior and prolonged safety violations. In this paper, we propose Constraint-Sensitive Policy Optimization (CSPO), a first-order primal-dual method that incorporates local constraint sensitivity into policy updates. CSPO augments the primal objective with a constraint-sensitive correction derived from the shortest signed distance to the safety boundary, enabling smarter recovery steps back to safety, compensating for delayed Lagrange multiplier updates, reducing oscillations near the boundary, and preserving the KKT solutions of the original constrained problem. Experiments on navigation and locomotion benchmarks demonstrate that CSPO achieves faster safety recovery and high reward preservation, resulting in higher constrained returns compared to state-of-the-art primal-dual and penalty-based methods

In this report, we present Hy-Embodied-0.5-VLA, abbreviated as HyVLA-0.5, an end-to-end system that spans the full robot learning stack: data collection, model design, continued pre-training and supervised fine-tuning, RL post-training, and real-world deployment. Each component serves a distinct role in this stack.

Recent advances in quantum computing pose a looming threat to most current Byzantine fault-tolerant (BFT) consensus protocols, which rely on quantum-vulnerable public-key signature schemes such as Ed25519 and BLS12-381. Instead of switching to much more expensive post-quantum secure signature schemes, an alternative is to use signature-free protocols, which rely only on cheap, post-quantum secure authenticated channels. In this paper, we ask whether signature-free BFT consensus protocols can match the performance of current state-of-the-art, quantum-vulnerable BFT consensus protocols. While previous work on the Sailfish++ protocol showed that state-of-the-art throughput is attainable signature-free, the question of latency is still open. Several recent signature-free protocols have low latency in theory, but they are all very intricate, and no practical implementation has so far been presented. In this work, we propose Simple-IT, a new leader-based, signature-free BFT consensus protocol that achieves a theoretical latency of 4 message delays (one more than the optimum), and only 3 on its optimistic path. Crucially, Simple-IT is simple enough to be amenable to implementation and to practical optimizations such as speculative pipelining, and, as we show experimentally in a geo-distributed testbed, it achieves both throughput and latency competitive with state-of-the-art quantum-vulnerable protocols.

Mixture-of-experts (MoE) layers enable the scaling of transformer models while keeping the inference compute fixed. While task-expert specialization has been observed in empirical studies of frontier MoE transformer models, existing theoretical work analyzes this using continuous mixture models that cannot be used to model natural language effectively. An important open question is to \textit{theoretically explain task-expert specialization in transformer MoE models using discrete models of language}. To address this, we represent structured knowledge via syntactic templates and finite key-value dictionaries, and prove formally that a single-layer MoE transformer can encode knowledge by using experts that specialize in the corresponding tasks. Our construction shows how queries are routed to unique, task-specific experts whose size depends solely on the intrinsic complexity of the given task (i.e. the combined size of its syntactic templates and factual dictionary). Our construction provides a theoretical support for empirical results on localized knowledge circuits in MoE models. We support our theoretical findings with experiments evaluating model performance under varying MoE loss functions.

As agentic systems continue to evolve and are widely deployed in real-world scenarios, there is a growing demand to faithfully evaluate their capabilities. However, current benchmarks are typically built on popular applications with relatively simple tasks and focus on a narrow set of capabilities while overlooking broader dimensions, resulting in saturated performance on modern agents and failing to probe their limitations. To this end, we introduce GauntletBench, a web-based benchmark for evaluating agent generalisation in challenging scenarios, focusing on three underexplored capabilities (temporal perception, graphical understanding, and 3D reasoning), across five less-covered professional applications (Video Editor, Workflow Builder, 3D Modeller, Flight Analyser, and Circuit Designer), each with 20 vision-intensive tasks (100 in total). Our benchmark provides a modular pipeline that comprises an environment compatible with both open- and closed-source agent frameworks, a controlled web-based application, a well-structured task suite, and an automated evaluation engine with diverse metrics. Contrary to widespread expectations, our empirical results reveal that frontier agentic systems remain far from achieving human-level performance. Even the state-of-the-art agent achieves only a 19.1% success rate on our GauntletBench, highlighting the limitations in these overlooked capabilities and generalisation. By comparison, non-expert human annotators achieve over 80% success on our challenging yet feasible tasks, revealing the substantial gap between current agent capabilities and those required for complex real-world scenarios.

Public blockchains continue to struggle with scalability because improving throughput is not as simple as increasing block size or reducing block interval. Larger blocks increase validation and transmission cost, while shorter intervals raise the likelihood of propagation delays, forks, and stale blocks. These limits motivate sharding, where transaction processing is divided across multiple parallel shard groups. In this work, we present a configurable SimPy-based discrete-event simulator for evaluating sharded blockchain architectures under controlled workload and network assumptions. The simulator models mining, verification, inter-shard coordination, block dissemination, measured throughput, average block time, and communication overhead. Our simulator achieves 1.6M TPS at 256 shards under a local datacenter-like setup and 0.6M TPS in a global WAN setup, showing strong throughput gains from parallel execution. However, the gains are not unbounded: beyond a certain number of shards, coordination traffic, synchronization, and network overhead begin to dominate, leading to diminishing returns.

Despite advances in large-scale Automatic Speech Recognition (ASR), disfluent speech remains challenging, as state-of-the-art systems are often optimized to omit disfluencies, leading to information loss and hallucinations. Prior work has focused on verbatim transcription and the integration of disfluency markers, but adapting models on limited datasets can lead to catastrophic forgetting of general-domain knowledge. We address this gap by leveraging continual learning (CL) with explicit disfluency tokens. We first introduce these tokens into a pretrained ASR model to establish stable token mechanisms, and then continue training on additional datasets with varying disfluency distributions. Through a detailed analysis of model dynamics during training, we identify a trade-off between marker learning and ASR performance, and a consistent cross-attention head mechanism shared across CL methods.

Interventions designed to modify a particular behavior in LLMs, such as refusal or sycophancy, often produce unintended changes in other behaviors. This lack of targeted control makes it difficult to design and implement reliable safety controls. To understand these side-effects, we introduce a diagnostic framework for analyzing interacting behaviors in LLMs. We model behaviors as low-rank subspaces in activation space, and study how interventions influence across behaviors. Across multiple instruction-tuned models (7B-70B) and across refusal, jailbreak, and sycophancy settings, we find that different behaviors share internal representations, and intervening on one behavior alters others in asymmetric ways. Some behaviors act as upstream control points whose interventions propagate broadly across other behaviors, while others remain more isolated. We relate these effects to two geometric quantities: (i) the overlap between behavior subspaces, measured as the average squared cosine of principal angles, and (ii) the angle between each behavior subspace and the decision subspace (capturing the model's final decision e.g., refuse vs. comply). Empirically, intervention effects on other behaviors tend to be larger for behavior pairs with higher subspace overlap, and for source behaviors whose subspaces lie closer (smaller angle) to the decision subspace. These findings highlight a challenge for targeted behavior control: behaviors are difficult to modify independently, as interventions can propagate through shared representations and asymmetric interactions.

Scientific discovery saturates when new hypotheses cease to provide independent information, even if the nominal hypothesis space remains large. We study hybrid discovery systems that combine structured local search with LLM-generated non-local proposals and pose the Search Compression Hypothesis: non-local exploration helps only when three geometric conditions co-occur: spectral compression, orthogonal escape from the explored span, and residual signal alignment with the target. We formalize these conditions, derive necessary conditions for hybrid advantage, and test the mechanism in controlled synthetic environments, large-scale A-share factor discovery, and symbolic-regression benchmarks; a public tabular operational sanity check tests the associated budget-allocation implication. Signal-planting and directed-versus-random experiments show that novelty alone is insufficient: random orthogonal jumps expand coverage but do not improve yield without predictive alignment. Across compression sweeps, real factor archives, and LLM-SRBench tasks, hybrid gains concentrate in weakly represented but target-bearing directions and vanish as the hypothesis space approaches full rank. The framework turns LLM-guided discovery from generic novelty search into a diagnostic procedure for deciding when directed non-local exploration is warranted.

Vision-language-action (VLA) models are powerful action generators for robot manipulation, but they are typically executed with fixed inference and replanning schedules. This rigidity ignores the uneven difficulty of robot control: contact-rich or uncertain states may need more computation and fresher feedback, while easier states can often be handled with fewer inference steps and longer open-loop execution. We propose Elastic Queries Reinforcement Learning (EQRL), a framework that makes each VLA policy query elastic. A lightweight latent-schedule adaptor jointly selects the latent input, denoising budget, and action chunk length, without fine-tuning the underlying VLA model. To make scheduling difficulty-aware, EQRL trains a critic over the joint latent-schedule action and derives a state difficulty signal from critic ensemble disagreement. This signal guides compute toward difficult states, while a learned residual allows task-driven correction. We formulate variable chunk execution as query-level macro-action RL with chunk-dependent discounting and an amortized number-of-function-evaluations (NFE) budget. Across simulation and real-robot manipulation, EQRL reduces amortized inference cost while preserving or improving task success.

The workflow from particle collision to physics analysis passes through a series of reconstruction steps that are traditionally modular and disconnected, with no shared representation linking low-level detector data to high-level analysis tasks. We show that casting event reconstruction as a machine learning problem naturally produces such a shared representation. We repurpose a machine learning model trained for particle-flow reconstruction (MLPF) to perform three distinct analysis tasks: jet flavor identification, jet energy regression, and missing momentum regression. By appending the per-particle latent representations learned during reconstruction as additional input features, we substantially improve over baselines that use kinematic features alone. We further demonstrate that a single linear layer trained using only the latent representations achieves competitive performance against state-of-the-art baseline architectures, and outperforms the baseline for missing momentum regression with approximately 35 times fewer parameters. These results demonstrate that the latent representations learned during reconstruction encode essential physics information needed for downstream analysis, establishing MLPF as a foundation model and offering a concrete step toward an end-to-end pipeline from detector data to physics analysis.

We study multi-domain LLM training in which two models, each stronger in a different domain, co-evolve by tutoring each other through on-policy feedback. Unlike one-way distillation or single-model fine-tuning, our goal is mutual Pareto improvement: each model improves across domains without losing its original strength. To this end, we propose On-Policy Co-Distillation (OPCoD), where each student's self-distillation is conditioned on its own correct rollout and feedback from its peer. To make feedback exchange effective, OPCoD uses cognizance-based gating to decide when to give feedback and feedback anchoring to ground feedback in the problem. On Science Q\&A tasks, OPCoD consistently outperforms baselines and achieves Pareto improvement across all evaluated domain pairs and students.

Machine learning (ML) pipelines require extensive data preparation, feature engineering, and integration across heterogeneous sources, making them tedious and error-prone to develop. While large language models (LLMs) have recently shown promise for assisting programming tasks, chat-based interfaces provide limited control over pipeline behavior and often produce code that is difficult to optimize or integrate into production systems. We demonstrate SemPipes, a novel programming model that extends ML pipelines with declarative, LLM-powered semantic data operators. SemPipes allows developers to specify high-level natural language instructions for data-centric operations, while seamlessly combining these operators with arbitrary Python code from standard data science libraries. For the semantic operators, it synthesizes specialized implementations at pipeline training time, conditioned on dataset characteristics and pipeline context, enabling the flexible yet controlled integration of LLM capabilities. We demonstrate SemPipes through SemPiper, an interactive interface that visualizes computational graphs of the pipelines, synthesized operator implementations, and optimization trajectories produced by an evolutionary search procedure. Attendees can explore three end-to-end scenarios, modify pipelines, inspect generated code, and observe how semantic operators are synthesized and iteratively optimized. The demonstration highlights how declarative semantic operators enable controllable, optimizable, and practical integration of LLMs into ML pipeline development.

Frontier coding models may spend substantial capacity learning not only program behavior, but also accidental entropy in human repositories. Such repositories contain valuable signals: tests, incidents, migrations, edge cases, product judgment, and operational history. These signals are entangled with framework churn, naming drift, generated-source ambiguity, dependency rituals, CI dialects, weak proof routes, and human-oriented review customs. We propose agent-first canonical code, a proof-carrying substrate that rewrites routine product software into canonical behavior profiles, typed change algebra, proof lanes, constrained edit grammars, semantic patch cells, runtime negative memory, and proof-carrying change objects. The core hypothesis is that quotienting software by behavior equivalence under a declared oracle can collapse equivalent encodings into governed representatives with explicit evidence and proof obligations. The endpoint is amortized cost per verified correct change, including source, context, reasoning, tools, verification, security, provenance, review, failed loops, defects, and foundry cost under a common oracle. Reported reduction bands are hypotheses, not measured frontier results. The proposed limit is a No-Accident Horizon: removable accident decreases until residual novelty, evidence, governance, risk, and future optionality dominate. For supported routine-product distributions, this gives a defensible planning target near 100-fold all-in cost reduction, not a guarantee for all software. Preliminary QLoRA experiments on Qwen2.5-Coder-14B show that 64,088 canonical trajectories are learnable and suppress tested forbidden-language markers, but do not establish behavior preservation, scaling economics, or verified-change cost. The contribution is a falsifiable program centered on minimum functional description length and verified-change cost.

Applications in the 3D Computing Continuum, which unifies edge, cloud, and space, require combining multiple AI tasks such as object detection, time-series analytics, and natural language processing into Compound AI systems. These systems must satisfy stringent Service Level Objectives (SLOs) on accuracy, latency, and cost. A key mechanism for maintaining SLO compliance of Compound AI systems is runtime model selection, where AI models are dynamically switched for each workflow task. However, existing distributed and compound AI frameworks do not natively support runtime model selection. We present PLAIground, a framework that enables runtime model selection for Compound AI systems. PLAIground introduces Compoundable AI Model (CAIM) abstraction, which decouples task semantics from AI model implementations via Task and Data Contracts, enabling model switching without workflow changes. Additionally, PLAIground introduces Pixie, an SLO-driven runtime model selection algorithm, which dynamically selects the most suitable model for each task during execution. Our evaluation on two realistic Compound AI workflows demonstrates that Pixie achieves up to 91.3% accuracy while maintaining SLO compliance where fixed-model strategies either violate cost and latency budgets up to 21x or miss accuracy targets by 4%.

Federated learning is well suited to edge environments but is often limited by the uplink cost of transmitting model updates. This Work-in-Progress paper presents MUFFLe, a communication-efficient update compression scheme that integrates generalized deduplication (GD) into the FedAvg pipeline. MUFFLe deduplicates repeated patterns across the update vector, yielding a fixed-rate, variable-count compression scheme. Preliminary experiments on IID MNIST with 20 clients show that MUFFLe reaches the target accuracy of $92.93\%$ with 38~MB cumulative uplink communication, compared with 75~MB for 8-bit quantization, 86~MB for Top-$k$ sparsification, and 310~MB for uncompressed FedAvg. These results demonstrate the feasibility of applying GD to communication-efficient federated learning.

Error-bounded lossy compression is a fundamental technique for managing the rapidly growing volumes of scientific data produced by modern simulations and observational instruments. Most state-of-the-art-compressors follow a prediction-residual paradigm, where compression effectiveness depends on the quality of the predictor: more accurate predictions generate smaller residuals that are easier to compress. This observation raises a question: can modern machine learning models serve as superior predictors for scientific data compression? Answering this question directly is challenging because developing compression-specific ML predictors requires substantial resources. Instead, we leverage the climate domain where highly accurate pretrained weather forecasting foundation models already exist, making them an ideal testbed. We present a framework that integrates spatial and temporal deep learning models into a conventional error-bounded compression pipeline. The framework supports auto-regressive forecasting models and avoids error accumulation. Using ERA5 climate data as a representative large-scale scientific dataset, we evaluate three distinct ML predictors: a VAEformer-based codec (CRA5), a graph neural network forecaster (GraphCast), and a vision-transformer forecaster (Aurora), against the state-of-the-art compressor SZ3.1 under identical quantization and entropy-coding backends. Our evaluation over approximately 1.7 TB of data reveals a surprising result: although ML predictors generate more accurate predictions and can improve reconstruction quality by up to 91% while achieving up to 9.6x higher compression ratios for highly predictable variables, they do not improve overall dataset-level compression ratio. We show that prediction accuracy alone is insufficient: the spatial structure of the resulting residuals plays a decisive role in entropy coding efficiency.

Artificial Intelligence (AI) systems must typically satisfy service-level objectives including accuracy, latency, and cost. The prevailing model-centric approaches select a monolithic model at design time and apply identical computation regardless of input difficulty, cannot decompose tasks across specialized components, and have knowledge that is fixed at training time. During runtime, this can lead to performance degradation and increasing costs. Because the model is the main design variable, it determines the majority of system behavior, coupling operational objectives to a single design-time choice. Addressing these limitations requires shifting from model-centric to system-centric design. Compound AI systems realize this shift by orchestrating multiple models, algorithms, and tools as distributed AI systems through explicit control logic. The performance of such systems depends on their workflow topology, the models assigned to each task, and the parameters governing runtime behavior. We present a design methodology that organizes this space along two dimensions, workflow topology and configuration selection, and identifies eight design patterns, each consolidating techniques to address a specific limitation of monolithic deployment. We validate our methodology through three case studies. Across our case studies, Compound AI configurations approach accuracy of monolithic models within 2.5 to 4 percentage points while reducing latency by up to 60% and cost by up to 71%. We show that model selection and parameter configuration jointly determine system performance, but the resulting design space grows combinatorially, as workflows compose more patterns and components. Thus, we identify five open challenges that define a roadmap from manually configured prototypes towards systems that automatically discover and maintain SLO-compliance in Compound and Distributed AI systems.

The increasing availability of large-scale textual corpora has opened new possibilities for data-driven, quantitative approaches to historical analysis using Natural Language Processing (NLP). However, diachronic corpora with historical relevance from the pre-digital era remain scarce and often incomplete. We present a quantitative approach to historical analysis based on the reconstruction and exploration of a diachronic corpus of around 600,000 articles from the Italian newspaper "La Repubblica", covering all the articles published from the 1st of January 1985 to the 31st of December 2000 - a period of major political, social, and geopolitical change in Italy and globally. Using NLP techniques, we analyze the text at both lexical and semantic levels; we then apply tools from complex systems and statistical physics to trace shifts in media discourse over time. This allows us to detect key transition periods, such as the transition from the First Republic to the Second Republic in Italy, or major international conflicts like the Gulf War or the Kosovo War, without relying on prior labeling. The results show how combining computational linguistics with ideas from complex systems can offer new quantitative insight into historical changes, opening up new paths for studying the dynamics of media and society through large-scale textual data.

Large language models exhibit strong multilingual capabilities, however, their internal representations are difficult to interpret. Understanding these interactions is important for ensuring reliable behavior in multilingual systems. Recent work has shown that causal-geometric structure can explain how certain concepts are encoded as approximately linear and separable directions, but whether this framework extends to multilingual models, where language identity is correlated and hierarchical, is underexplored. We apply causal-geometric analysis to multilingual LLMs, studying 28 bilingual contrasts across three models, allowing us to analyze when languages behave as approximately independent factors and when structured dependencies persist. We find evidence that language concepts admit stable linear representations that are largely separable under a covariance-adjusted (causal) inner product, with structured deviations reflecting linguistic similarity. Moreover, languages within the same family (such as Germanic or Romance) exhibit a simplex-like geometric structure, suggesting hierarchical organization. These results extend causal-geometric interpretability to multilingual settings and provide insight into how separability and similarity may exist in multilingual LLM representations, motivating interpretability analyses that diagnose when and how structured dependencies between concepts can be anticipated. This has implications for trustworthy deployment, as residual structure between languages may lead to unintended cross-lingual effects when models are monitored or intervened upon.

Unstructured pruning produces sparse weight tensors, but the standard implementation keeps tensor shapes unchanged so the deployed model is no smaller than before pruning. We present an exact structural rewrite, which we call minimization, that converts a masked network into a smaller dense network with the same forward function up to floating-point rounding. The Squeeze-Release cycle iterates pruning and minimization with an intermediate release step that re-enables the exact-zero positions inside the compacted tensors as small calibrated noise, turning otherwise wasted capacity back into trainable parameters. Successive cycles use that capacity to find structural redundancy a single pass cannot reach. We additionally introduce CompensatedLayerNorm, a function-preserving replacement for LayerNorm that extends minimization to channel reduction across LayerNorm-equipped residual streams. Squeeze-Release compresses the deployable network to 39x smaller than the unpruned model on a fully-connected model network and 14.8x smaller on modern CNN (ConvNeXt-Tiny), at comparable accuracy. In addition we prove that the rewrite can be extended to transformer architectures.

Tactile imaging seeks to reconstruct the internal structure of soft objects through touch sensing, with applications in medical diagnosis and robotic manipulation. Recent self-supervised learning approaches have shown promising results, but rely on global, unstructured representations and robot-controlled sensing, limiting generalization and practical use. We propose Local Encoder for Spatial Sensing (LESS), an object-centric tactile representation that exploits the local nature of touch. The tactile scene is modeled as a grid of recurrent encoders with local receptive fields, whose states are fused to reconstruct 2D or 3D images of internal structure. This compositional design enables strong generalization: models trained on single-inclusion phantoms accurately image objects with multiple inclusions and varying sizes. The local structure further supports spatial uncertainty estimation. In addition, we enable hand-held tactile imaging via external pose tracking and human-like palpation data, and extend tactile imaging to full 3D reconstruction.

Recovering structural information from noisy high-dimensional data is a fundamental task in statistical inference. We investigate the recovery thresholds for a graph hidden in a randomly weighted complete graph. Specifically, an unknown graph $H^* \in H_n$ is chosen uniformly at random, and hidden in a complete graph of $n$ vertices as follows: the weight of an edge $e \in H$ is distributed independently according to $P_n$; otherwise the weight is distributed independently according to $Q_n$. The goal is to recover almost all of $H$ from these edge weights. Assuming a local Lipschitzness of the Rényi divergence between distributions $P_n$ and $Q_n$, and a mild density condition for the graphs $H_n$, we give a unified characterization of the information-theoretic limit for recovering almost all of $H$ (also known as almost exact recovery). Our characterization connects the KL divergence between $P_n$ and $Q_n$ to the logarithm of the first moment threshold of $H$ in the Erdős-Rényi random graph model $G(n,p)$. Our lower bound also extends to the task of partial recovery, in which only a constant $λ$-fraction of $H$ needs to be recovered. Last but not least, for certain Bernoulli and Exponential regimes, and for Gaussian distributions, we are able to show an All-or-Nothing (AoN) threshold phenomenon at the exponential scale.

High-dimensional datasets often concentrate near low-dimensional structures, but estimating their geometry from samples typically relies on graphs and kernels that scale poorly with dataset size and dimension. We propose Riemannian metric matching: a denoising probabilistic framework for learning the Riemannian geometry of data using neural networks. Specifically, we learn the carré du champ operator, which, using diffusion geometry, gives us access to the Riemannian geometry toolkit for downstream machine learning and statistical tasks. Our key observation is that the carré du champ operator can be formulated as a conditional expectation over random perturbations of the data, which can be exploited for sample-wise training and constant cost, amortized inference without explicit kernel construction. Empirically, metric matching rivals or improves the accuracy of $k$-NN-based diffusion geometry estimators, while enabling amortized inference that is up to $400\times$ faster, and supports graph-free geometric analysis on high-dimensional images where nearest neighbors break down.

This paper appraises recent frameworks within AI development to integrate LLMs into control tasks in automotive contexts from the perspective of safety assurance. This work has built upon the rapid integration of LLMs across automotive settings. However, we find that at present, these frameworks face significant challenges, limiting their efficacy in real-time safety-critical contexts. Firstly, we consider conceptual challenges, including the fact that deployers are faced with a dual challenge, wherein they must assure a model which has been developed upstream, i.e. as general-purpose tools by the large AI labs, in a downstream context, i.e. into specific vehicle architectures. Secondly, we consider concrete challenges from across existing standards. We show that there are currently both fundamental engineering constraints covered in ISO21448, such as latency, and novel LLM-specific issues, such as alignment-related issues covered in ISO/PAS8800. We ground both examples in a concrete introductory, experimental case study exploring an existing open-source repository, Talk2Drive. We present a safety argument in order to make explicit the limitations of existing solutions. Nonetheless, given that the use of LLMs in automotive contexts is being explored at a technical level and operationalised, we propose potential assurance mechanisms for LLM-related hazardous events going forward.

Property Graphs are rapidly being adopted as database frameworks for representing heterogeneous data sources. To enable precise access to the information contained in them we need conversational interfaces based on Text-To-Cypher (Text2Cypher) parsers. This paper presents an automatic synthetic data generation method that can be leveraged to fine-tune small LLMs for this task. We conduct experiments on all the major Text-To-Cypher benchmarks, demonstrating that with our synthetic data generation approach we can significantly increase the performance of small LLMs, allowing them to compete with much larger proprietary models. This means that in settings in which models must be locally deployed we can ensure data-sovereignty without sacrificing accuracy and without costly annotation campaigns.

LLM agents have rapidly evolved into autonomous systems, yet a persistent information gap remains between users and agents: communication is costly, while users' identical preferences further limit information exchange. To investigate how agents should communicate across modalities, this paper formalizes Communication Policy, establishes textual and UI-based policies, and then evaluates communication policies across diverse environments, personas, and model combinations. Building information asymmetry for proactive agents, we set up two complementary settings, User-Agent and Planner-Executor. Experimental results reveal complementary strengths between interaction channels: text-based interaction often facilitates task performance, while structured UI improves agents' response quality and persona compliance. Motivated by that, a hybrid method combines these advantages. We further propose Communication Policy Evolution (CPE), a self-evolution framework for refining communication policies through rollout and prompt-level evolving. Without model modification, CPE achieves the best task success across multiple settings using prompt refinement alone. Our findings identify communication behavior as a critical yet underexplored design dimension for LLM agents.

Real-world spatio-temporal forecasting must handle irregular time points, spatially sparse observations, and the need for uncertainty quantification. This setting is often further compounded by nonlocal interactions (long-range spatial coupling). Modeling continuous-space, continuous-time nonlocal dynamics naturally leads to infinite-dimensional integro-differential equations (IDEs), making principled Bayesian inference intractable. We propose the NonLocal Bayesian Spatio-Temporal model (NLBST), a hierarchical Bayesian framework for continuous spatio-temporal fields that learns explicit nonlocal coupling while retaining tractable inference. NLBST represents the latent field via a coordinate-based spatial basis expansion and models the coefficient process with a continuous-time ODE whose learnable linear operator corresponds to a Galerkin reduction of a nonlocal IDE; a Neural ODE residual captures additional nonlinear dynamics. A linear-Gaussian observation model enables Kalman-style sequential updates under missing and irregular observations, while the spatial basis representation enables inductive prediction at unmeasured locations without retraining. Global parameters are learned via variational inference, and uncertainty is handled through a Bayesian hierarchy. Experiments on synthetic and real-world datasets demonstrate strong forecasting and spatial generalization with well-calibrated uncertainty, yielding substantial gains over baselines in strongly nonlocal and partially observed regimes.

Property graphs may be constrained by schemas that inform both query engines and human users about the shape of valid data, enforcing a contract between data provider and consumer. Composable property-graph queries transform input graphs into output graphs. Then, the question arises of which schema can be expected after one (or several) transformation steps. We investigate how schema constraints can be inferred given an input schema and a transforming query. Specifically, we propose a reasoning procedure that, given an input schema in ProGS and a query in G-CORE infers an output schema. Since graph updates will happen frequently, our inference procedure does not rely on graph instances, such that the computed output schema applies to all graphs originating from any input graph complying with the input schema. Related work has addressed this problem for SPARQL CONSTRUCT queries, encoding it in Description Logics (DLs) so that the output schema is entailed by axioms inferred from input schema and queries. Property graphs and their queries, however, complicate the matter, as property graphs feature label and property annotations as well as first-class edges. Thus, reification has to be used in one way or another, though available DLs lack the means to encode such features directly. We approach this novel challenge via a family of mappings for i) property graphs reified in RDF, aligned with ii) a mapping from ProGS to SHACL and iii) a mapping from G-CORE to SPARQL CONSTRUCT queries. In this manner, schema inference for property graphs becomes manageable, as we break apart the problem through the extra mapping layer and utilize efficient DL reasoners. We develop the metatheory regarding the soundness of inferred schema constraints and the semantic equivalence of mapped schemas and queries.

Current Generative AI (GenAI) interfaces remain largely constrained to chatbox interaction, which can impose high cognitive demands on users and create substantial barriers for people with intellectual disabilities (ID), including prompt formulation difficulties, response overload, and limited mechanisms to assess information reliability. To explore alternative interaction models for cognitive accessibility, we conducted a cross-disciplinary co-design challenge in which two student cohorts (Computer Science and Industrial Design) developed interface concepts from the same set of functional requirements (e.g., prompt scaffolding, structured output, GUI-based refinement, transparency, and personalization). Comparing the resulting proposals reveals both convergence on foundational requirements (notably initial calibration, proactive prompting, and direct manipulation of response fragments) and complementary contributions that outline a multi-layered support system. Computer Science teams primarily produced structural scaffolding, emphasizing predictability, navigability, and trust through mechanisms such as reliability indicators, explicit sources, and context management for long conversations. Industrial Design teams emphasized experiential scaffolding, focusing on pacing, attention guidance, multimodality, and proactive agency, including step-by-step response flows, focus modes, and assistant-like integrations. We synthesize these findings into a dual-layer scaffolding framework that expands the design space for cognitively accessible GenAI interaction beyond chat-centric models and motivates future work on expert refinement, technical feasibility, and empirical validation with users with ID.

LLM-based agents trained with reinforcement learning optimize step-wise action prediction but lack metacognitive awareness of task progress, inducing a gap that hinders long-horizon scaling. A pilot study reveals that online progress prompting hurts performance while retrospective demonstrations help, yet this capability cannot emerge from outcome-reward training alone. We present RePro, Retrospective Progress-Aware Training, a framework that trains agents to self-generate progress signals via a forward-then-reflect rollout paradigm: the agent executes actions online, then retrospectively reassesses its step-wise progress given the completed trajectory and known outcome. RePro initializes with a Retrospection Warmup that teaches reflection format from minimal external demonstrations, then further trains through RePro-PO with a composite reward that produces self-generated signals without continuous external supervision. Experiments on WebShop, ALFWorld, and Sokoban show that RePro enhances the Qwen family's performance, with up to $12\%$ absolute success rate gains.

Vision-Language Models (VLMs) such as CLIP have become a standard backbone for open-vocabulary recognition, yet their zero-shot predictions remain vulnerable to distribution shifts encountered at deployment. Test-Time Adaptation (TTA) has recently been extended to CLIP as a lightweight solution, leading to a rapidly growing body of TTA4CLIP methods. However, empirical progress in this area has largely outpaced our understanding of what truly drives adaptation, where their gains originate, and under which shifts they remain reliable. In this paper, we take a step back from the pursuit of state-of-the-art accuracy and conduct a systematic controlled study of TTA4CLIP. We first organize existing methods into three unified paradigms according to what is updated at test time. We then introduce TTABC, an open-source TTA Benchmark for CLIP, which standardizes evaluation protocols and integrates more than 20 representative methods. Our controlled empirical analysis focuses on three key areas. First, we determine the driving factors in parameter-based methods, revealing that adaptation gains are primarily driven by test-time evidence and reliable proxies rather than heavy optimization. Second, we explore evidence utilization beyond heavy parameter tuning, showing that competitive and efficient performance can be achieved through cross- or current-sample evidence and lightweight prototype updates. Finally, we demonstrate that there is no silver bullet for TTA: no single adaptation paradigm is universally optimal, and the preferred paradigm depends on the nature of shift. We hope our benchmark and study provide a clearer understanding of the current TTA4CLIP landscape and establish a foundation for further research.

Developing accurate crowd-counting models for Hajj pilgrimage scenes remains challenging because domain-specific annotated images are scarce and data collection during large gatherings raises privacy concerns. To address these limitations, this paper proposes Pix2Pix-Hybrid (P2P-H), a hybrid conditional GAN for structure-guided Hajj crowd-image synthesis and data augmentation. P2P-H builds on Pix2Pix and employs a U-Net generator conditioned on eight input channels that jointly encode structural cues (edges and grayscale) and contextual attributes (crowd density and time of day). To capture detailed textures in dense scenes, the framework integrates two multi-scale PatchGAN discriminators operating at different resolutions. The training procedure combines adversarial, perceptual, and feature-matching objectives with adaptive data augmentation and stabilization strategies. The model was trained on 993 real Hajj frames collected from 60 publicly available video sources, with conditioning attributes derived automatically to reduce manual labeling effort. Using this framework, we constructed CrowdH, a synthetic dataset of 10,000 high-resolution Hajj crowd images. Experimental results show that P2P-H improves structure-preserving conditional synthesis quality compared with Pix2Pix and StyleGAN2-ADA baselines and shows favorable transfer to other crowd datasets. To assess downstream utility, we further constructed CrowdH-Mix-469, an annotated mixed real-synthetic dataset comprising 384 real Hajj images and 85 selected synthetic images,and evaluated five crowd-counting models under real-only and real-plus-synthetic training. The selected synthetic data reduced MAE across all five models, with the strongest gain observed for CSRNet.

Frontier AI systems are increasingly capable of cybersecurity tasks, including codebase inspection, vulnerability detection, and exploitation. However, evaluating their offensive capabilities remains constrained by limited access to open, reproducible, multi-host cyber ranges. Existing public benchmarks capture isolated skills such as CTF solving, vulnerability reproduction, and exploit generation, but often abstract away realistic intrusion workflows: discovering exposed services, gaining a foothold, collecting internal information, and expanding compromise across hosts. This gap makes it difficult to observe emerging risks early, because frontier AI systems are rarely evaluated under realistic attack conditions. We introduce AgentCyberRange, the first open, multi-range infrastructure for measuring autonomous cyber attack capability in realistic cyber ranges. It combines 110 vulnerabilities across 15 real web applications and 8 enterprise-like cyber ranges with 156 internal hosts, plus Cage, a toolchain for execution, orchestration, result collection, and verification. The benchmark covers two core stages: web exploitation, where agents explore exposed applications and validate vulnerabilities, and post exploitation, where agents turn an initial foothold into broader internal compromise. We evaluate six frontier AI systems under matched prompts and budgets. GPT-5.5 with Codex performs best, solving 16.1% of web exploitation tasks and 31.7% of post-exploitation tasks; with more concrete hints, these rates increase to 33.0% and 46.3%. We also observe out-of-benchmark findings, including unknown vulnerabilities in popular projects, and payload mutation that bypasses host defenses. These results show that open cyber-range evaluation is necessary for observing emerging offensive capabilities under realistic and reproducible conditions.

Population-based and distributional optimization methods, from evolution strategies and consensus-based optimization to covariance-matrix adaptation and stochastic gradient methods viewed as distributional dynamics, are widely used for nonconvex or black-box problems, yet their convergence analyses remain fragmented across algorithm-specific techniques. We introduce an operator calculus in which a broad class of such methods, after choosing an appropriate state space and, where necessary, augmenting the state by memory or strategy variables, is described as a composition of three elementary operators (mutation, selection, and recombination) acting on probability measures. Under explicit stability and regularity conditions, the composite operator admits a pre-generator whose continuous-time limit is a transport-reaction-jump (TRJ) PDE that preserves the operator splitting. On this foundation we establish a modular Lyapunov principle. If a state-space Lyapunov function both dissipates under the full generator and controls the relevant search-space gauges, then the state-space Lyapunov functional and the induced search errors decay exponentially. The additive generator structure allows dissipation estimates to be assembled operator by operator, providing a toolkit for certifying convergence of composite mean-field algorithms.

Time series prototype learning is fundamentally challenged by observational ambiguity. Discrete architectures fail to resolve this, as they lack the capacity to decouple stochastic noise from continuous dynamics. Furthermore, rigid closed-set assumptions fail to capture unseen diversity. To address these limitations, we propose a hierarchical ordinary differential equation clustering network, which utilizes neural ordinary differential equation to model latent state evolution as a continuous integral curve. This formulation enforces temporal continuity to effectively disentangle smooth feature trends from stochastic noise, while our adaptive hierarchical mechanism autonomously determines the appropriate number of prototypes without rigid prior constraints. Validated on the early link failure detection task with irregularly sampled time series, the proposed method effectively extracts underlying physical prototypes, thereby enabling robust failure detection. Our code is available at https://github.com/NJ-LNN/Hierarchical-ODE.

Deep learning has driven many recent advances in process analytics, especially for predictive and prescriptive monitoring. However, standard objectives such as cross-entropy optimize local next-step likelihoods and only implicitly capture control-flow structure. As a result, models can achieve high token-level accuracy while permitting imprecise global behaviour. We introduce DIFF-ERO, a conformance-aware loss function for deep learning models on process data. DIFF-ERO is a differentiable formulation of entropy-based stochastic conformance that incorporates control-flow information during training. Our approach constructs batch-level stochastic transition matrices with soft edge memberships, allowing structural precision and recall signals to directly inform backpropagation. The loss is model-agnostic and can be applied whenever the final representation parametrizes stochastic transitions. We instantiate DIFF-ERO in transformer encoder-decoder pipelines for next-activity prediction and use it jointly with cross-entropy to analyse its theoretical components with respect to convergence. Across benchmarks comparing other loss functions and targets, DIFF-ERO shows improved predictive performance where structure matters most while maintaining parity elsewhere. At the same time, the learned stochastic automaton converges towards the structural ground truth, indicating that the network internalizes process model structure.

Large language models (LLMs) are now widely used as automatic judges for open-ended instruction-following evaluation. This practice is convenient, scalable, and often more semantically aware than reference-based metrics, but it also introduces a new reliability question: does a judge evaluate the quality of an answer, or does it also react to the language in which the comparison is presented? We propose Judge-LS, a lightweight meta-evaluation protocol that transforms LLMBar response-pair items into English, Chinese, and Chinese-English language-switched variants. A reliable judge should preserve its preference under label-preserving language transformations and should not prefer a language when two answers are translation-equivalent. We evaluate four API-accessible judges on the full 419-item LLMBar benchmark, producing 13,408 successful pairwise judgments. Across models, Chinese and language-switched presentations induce 10.7--14.4% preference flips relative to English, and all judges achieve their highest accuracy in English. However, translation-equivalent tie probes do not reveal a systematic English preference: most probes are judged as ties, and non-tie decisions more often favor Chinese. We add confidence intervals, paired significance tests, and an automatic transformation audit with a sensitivity analysis that excludes mechanically flagged high-risk variants. The experiment requires no model training, uses only API calls, and is feasible on modest local hardware.

Fall recovery is critical for autonomous legged locomotion. Existing methods have demonstrated that some legged robots, such as humanoids and quadrupeds, are capable of fall recovery from diverse postures by utilizing arms or coordinating multi-legs to generate support forces. Without arms or other legs to provide supportive assistance, a bipedal-wheeled robot must rely solely on the actuation of its legs, making recovery particularly difficult. To address this, we introduce FTSR (Force-guided Teacher-student framework with Stage-wise Rewards). The force-guided method constructs an external auxiliary force during simulation training that correlates directly with the robot's real-time height, explicitly formulating this force as an optimizable constraint. Through constrained reinforcement learning, the policy is guided toward reducing force dependency gradually and increasing the body height, developing internal recovery strategies despite having no arms for support. Height-progressive stage-Wise rewards progressively structure posture stabilization during recovery and transition to sustained locomotion, integrated with teacher-student architecture distilling privileged knowledge of force effects and recovery dynamics. After simulation training, the policy is deployed on a physical armless bipedal-wheeled robot and extensively evaluated. Experiments confirm robust and reliable fall recovery under diverse challenging conditions, demonstrating strong environmental adaptability and motion robustness, while maintaining full post-recovery motion capability. The framework also generalizes effectively to a high-DOF humanoid, confirming its practical generalizability. The project page is available at https://2350575870.github.io/force-guided.github.io/

Fixed-cardinality retrieval injects a constant top-K chunks into the generator regardless of query complexity, causing over-retrieval for narrow queries and under-retrieval for compositional ones. We describe ScoreGate, a lightweight score-space decision mechanism that controls retrieval cardinality at inference time using two scores already produced by the standard pipeline: bi-encoder similarity s_i and cross-encoder reranker score r_i, with no additional model inference calls required. Its core insight is that cross-encoder affirmation can rescue semantically relevant chunks that bi-encoder retrieval ranks poorly due to vocabulary mismatch -- a failure mode unaddressed by fixed-K or single-score thresholding. On MS MARCO (200 dev queries), ScoreGate achieves MRR@10 = 0.401 with 35% fewer retained chunks than Standard Top-K. On an internal benchmark (n=300, Fleiss' kappa=0.87), ScoreGate observed zero false positives (95% CI [96.4%, 100%]) at 97.77-99.34% recall, with 34.8% fewer tokens per query and only 31ms added latency. Results on both MS MARCO and real-world production traffic suggest that adaptive retrieval cardinality can improve retrieval efficiency without degrading retrieval quality.

We develop a statistical learning theory for gradient boosting applied to the estimation of covariate-dependent Generalized Pareto (GP) distributions in the context of Peaks-over-Threshold modeling. After an orthogonal reparametrization of the GP likelihood that diagonalizes its Fisher information matrix, we cast the estimation problem within the Empirical Risk Minimization (ERM) framework and derive non-asymptotic error bounds for the boosting estimator. Our analysis accounts for three distinct sources of error in the process: statistical fluctuations, the approximation bias inherent to the asymptotic nature of the GP model-controlled under second-order regular variation-and the approximation error associated with the finite number of boosting iterates, making explicit the resulting bias-variance trade-off. We illustrate the practical benefits of the reparametrization through simulations, showing that it significantly reduces gradient correlation during training and improves convergence stability. The methodology is applied to a medical malpractice insurance dataset from the Texas Department of Insurance, comprising over 18 000 closed claims. The gradient boosting approach yields a good fit for the tail of settlement cost distributions and reveals that the number of days to settlement is the dominant predictor of tail heaviness, consistent with earlier findings in the reserving literature.

Secure software engineering in practice is a multi-stage workflow involving vulnerability analysis, remediation, and fix verification. However, current LLM-based software security approaches often focus on isolated tasks such as detection or patch generation, with limited attention to agentic architectures reflecting industrial workflow. This creates a gap between existing LLM-based vulnerability-handling methods and real-world practices. In this paper, we study a role-based agentic workflow for vulnerability analysis and mitigation consisting of Planner, Analyzer, Fixer, and Verifier roles. To explore the effect of static analysis tool, the analyzer agent was integrated with the CodeQL in one of the workflows. The models used include nemotron-cascade-2:30b, qwen3-coder-next, and gpt-oss:120b. Our evaluation uses 25 real-world C/C++ vulnerabilities. The study reports 44% vulnerability detection accuracy comparable to GPT 5.5 and 19% fix accuracy. We also list implications from this study in context of software security practitioners.

Semantic IDs are crucial in generative recommendation, but with a fundamental limitation: temporal information is not well incorporated into semantic IDs. Instead, time influences recommendation only implicitly (e.g., through session construction heuristics, preference alignment, or sequence order), while existing semantic ID learning remains entirely time-agnostic. This design conflates interactions occurring under distinct temporal contexts into identical semantic representations, implicitly assuming that item semantics and user intent are temporally stationary. Such an assumption is misaligned with real-world recommendation scenarios, where evolving interaction rhythms play a central role. In this work, we investigate where and how the explicit time should be incorporated into semantic ID for generative recommendation. First, we systematically characterize the design space along three orthogonal dimensions of temporal signals and present a unified framework, ChronoID, for time-aware semantic ID learning. Then, by contributing a new time-explicit generation recommendation benchmark, ChronoID answers the questions: what is the effective way of infusing time, how to design the architecture, and where does the gain come from.

Prior work has identified several factors that can contribute to the performance gap between Adam and SGD, spanning data aspects, architecture design, and optimization properties. Yet these explanations are often studied in isolation, leaving their relative importance unclear. In this work, we revisit these hypotheses through a controlled empirical study across vision, language, genomics, and graph tasks, spanning modern and classical architectures, and carefully designed training setups. Our results suggest that no single factor consistently explains the Adam--SGD gap. For instance, the Adam advantage can (1) persist under a uniform vocabulary distribution yet nearly disappear under a heavy-tailed one; (2) reverse in favor of SGD in softmax-attention models; and (3) become larger under soft architectural modifications, e.g., when ReLU is replaced by a GeLU nonlinearity. This suggests that the gap arises from nontrivial data and architecture interactions, rather than from a single common factor. Yet, we observe a pattern across our settings: a \emph{crossover batch size} at which the relative advantage shifts from SGD to Adam as the batch size scales. These empirical results are captured by our theoretical gap model, which predicts this batch-size-dependent crossover. Our perspective helps reconcile several existing hypotheses while offering practical insights across domains.

Linguistics olympiad problems (LOPs) are a category of self-sufficient puzzles consisting of a scaled-down corpus representative of certain linguistic phenomena, from which the solver must deduce a primitive set of rules of the language and then translate a new set of elements. The linguistics olympiads (LOs) have become a worldwide phenomenon with 43 different territories taking part in the International Linguistics Olympiad (IOL) 2025. While the typology and solving strategies of LOPs have been analysed, their scientific facet and connections to academic linguistics have yet to be explored. LOPs are directly connected to many linguistic fields, e.g., linguistic typology, linguistic relativity, and linguistics fieldwork. Recently, LOPs have become a research focus as benchmarks for large language models, thus highlighting their usefulness in computational linguistics. Nevertheless, they have not yet been integrated into mainstream linguistics research. This paper attempts to open new directions of including this particular type of puzzle in academic research by offering a structured evaluation of LOPs as linguistic data sources and proposes criteria for their responsible use in academic research. Starting from a set of over 1800 LOPs, this study critically examines the potential of LOPs as a novel corpus for linguistics research by discussing their strengths and limitations as tools, as well as the areas of linguistics into which these problems could fit. This work forms the foundation for a broader initiative aimed at bridging the gap between LOs and academic linguistics, by establishing a robust theoretical framework for LOPs.

AI agent performance depends critically on the runtime harness, comprising the prompts, tools, memory, and control flow that mediate how a model observes, reasons, and acts. Yet today's harnesses remain largely hand-crafted and static: each new model or task still demands bespoke scaffolding, and the rich traces produced during execution are rarely distilled back into systematic improvement. We introduce HarnessX, a foundry for composable, adaptive, and evolvable agent harnesses. HarnessX assembles typed harness primitives via a substitution algebra, adapts them through AEGIS, a trace-driven multi-agent evolution engine grounded in an operational mirror between symbolic adaptation and reinforcement learning, and closes the harness-model loop by turning trajectories into both harness updates and model training signal. Across five benchmarks (ALFWorld, GAIA, WebShop, tau^3-Bench, and SWE-bench Verified), HarnessX yields an average gain of +14.5% (up to +44.0%), with gains largest where baselines are lowest. These results suggest that agent progress need not come from model scaling alone: composing and evolving runtime interfaces from execution feedback is an actionable and complementary lever. The complete codebase will be open-sourced in a future release.

Drug-target interaction (DTI) and affinity (DTA) predictors increasingly achieve strong benchmark scores, yet their internal use of sequence, fingerprint, and graph features often remains opaque. We present an interpretability audit of BridgeDPI architecture on three different datasets including Gao, Human, and C.elegans. This study combines gradient-based attributions -- integrated gradients, saliency, layer-wise relevance propagation, SmoothGrad, and SmoothGrad-IG -- with feature-wise occlusion ablation and strict intersection consensus across methods to reduce single-explainer bias. We summarize sensitivity and signed effects at raw inputs, at the bridge similarity scaffold, and through the graph convolution, including edge-level sensitivities and targeted edge removals. The results show that explainability is most informative when treated as model criticism: it reveals modality dominance, padding and special-token artifacts, dataset-dependent cooperative versus suppressive effects across layers, and chemistry-consistent fragment and composition motifs where methods agree. These analyses do not substitute for structural or experimental ground truth, yet they can provide testable hypotheses for downstream validation in computational drug discovery pipelines. More broadly, applying modern XAI to contemporary DTI/DTA models is still an early pass over the rich structure implicit in trained weights and data -- yet even this first layer of scrutiny already helps researchers relate predictions to drug- and target-side representations and to prioritize external validation.

Knowledge injection aims to equip large language models (LLMs) with external, domain-specific, or time-sensitive knowledge. Existing approaches typically face a trade-off between flexibility and integration: retrieval-augmented generation keeps knowledge outside the model but only provides prompt-level augmentation, whereas post-training based methods encode new knowledge into shared parameters but may introduce catastrophic forgetting, knowledge conflict, and costly updates. In this paper, we propose Decoupled Mixture-of-Experts (DMoE), a modular architecture for parametric knowledge injection that decouples both experts and the router from the base model. DMoE converts external knowledge corpora into independently updatable expert modules and uses a lightweight uncertainty-aware router to activate relevant experts only when the base model lacks sufficient knowledge during generation. To support efficient auto-regressive inference, DMoE attaches experts only to the final-layer feed-forward network, preserving KV-cache reuse while enabling parameter-level knowledge augmentation. Experiments on knowledge-intensive benchmarks show that DMoE consistently improves answer quality over retrieval and adapter-based baselines.

Affordance reasoning, the inference of an object's action possibilities from its physical properties (e.g., shape and material), is fundamental to human physical understanding and increasingly critical for Large Language Models (LLMs). However, existing affordance benchmarks largely expose explicit object identities in the evaluation setup, allowing models to rely on memorized object-affordance mappings rather than reasoning over physical properties. To address this gap, we introduce Affordance20Q, a novel affordance reasoning benchmark formulated as a 20-Questions game without exposing the object's identity. In each game, the model identifies a hidden object's affordance from a candidate set by asking yes/no questions about its physical properties. Affordance20Q comprises 1,009 games over 454 objects and 59 affordances, all manually filtered, refined, and annotated. We conduct comprehensive experiments with 15 state-of-the-art LLMs and find a substantial gap (~20 points) compared to human performance. A KL-based information-gain (IG) analysis further shows that models fail to ask discriminating questions as the game progresses. To close the gap, we develop KB-Anchored Rule Induction (KARI), a pipeline based on LLMs that generates affordance rules grounded in evidence from knowledge bases (KBs). KARI improves open-source LLMs by up to 15.2 points, while the limited coverage of KBs hinders further gains. We release all our code and data at https://github.com/1171-jpg/Affordance20Q.git

Agent skills are structured procedural packages that guide frozen LLM agents in specialized workflows. Skills rarely remain sufficient after deployment: edge cases, API changes, and deployment constraints become visible only through use, making skill evolution a practical necessity. Existing methods depend on privileged feedback such as held-out validation scores, hidden test outcomes, or environment rewards -- signals often unavailable when a practitioner has only a task description and workspace data. We introduce SkillAudit, a framework for evolving agent skills without ground-truth feedback. The key idea is paired trajectory auditing: at each iteration, the same task is executed with and without the candidate skill, isolating how the skill changes agent behavior without external labels. To turn behavioral differences into edit guidance, SkillAudit uses Process-Aligned Contrastive Evaluation (PACE), a cluster of evaluators that maps trajectory divergences to diagnostic signals linked to specific passages in the skill document. A structural verifier, compiled once from the task specification and then fixed, checks task constraints and rolls back harmful updates. SkillAudit routes edits through two pipelines: Refine removes noisy or irrelevant guidance from broadly useful skills, while Repair replaces passages that conflict with the task. Across 89 containerized tasks spanning 8 professional domains, SkillAudit achieves 73.9% average task reward, outperforming an agent without skills (40.9%) and the static expert skill (56.7%). These gains are obtained without accessing hidden tests, reference solutions, or external scoring functions during evolution.

Safety certification of Vision-Language-Action (VLA) driving planners under ISO 21448 (SOTIF) rests on an Operational Design Domain (ODD) specification that answers two complementary questions: when does the planner start to fail, and how severely does it fail once it does? We evaluate Alpamayo R1, a 10B-parameter open-weight driving VLA, on 15,968 (clip, attack) pairs. We find a conservative-aggregate gap: an aggregate safe threshold of $σ\leq 50$ under a 15% average displacement error (ADE) budget masks well-sampled scenarios that tolerate the top of the tested grid ($σ= 70$). A Gaussian Mixture Model (GMM) on the changed-explanation subset identifies six discrete severity bands (BIC-optimal $k{=}6$), so two perturbation conditions with the same mean error can differ materially in their share of high-severity (C4/C5) failures. Joining the two analyses on the same corpus surfaces a finding neither yields in isolation: the scenarios with the loosest noise thresholds are not those with the lowest high-severity rate: STOP_SIGNAL concentrates roughly $4\times$ the C4/C5 share of LANE_KEEPING despite tolerating a larger $σ$. A deployable SOTIF ODD specification for driving VLAs therefore requires a two-dimensional safety envelope, not a single aggregate value per hazard.

Variational Inference (VI) is a fundamental inference technique in Bayesian machine learning for approximating complex posterior distributions. Traditional VI often relies on the mean-field factorization, which can inadequately capture true posterior complexity. Recent advancements have leveraged neural networks to model implicit distributions, offering increased flexibility. However, the practical constraints of neural network architectures still produces inaccuracies. In this paper, we propose a method called Implicit Variational Rejection Sampling (IVRS), which integrates implicit distributions with rejection sampling to improve the posterior approximation. Our method uses neural networks to construct implicit proposal distributions, and rejection sampling with a discriminator network that estimates the density ratio between the implicit proposal and the true posterior for refining the approximation. Towards this end, we introduce the Implicit Resampling Evidence Lower Bound (IR-ELBO) as a metric to characterize the resampled distribution's quality and derive a tighter variational lower bound. Experimental results demonstrate that our method outperforms traditional variational inference techniques.

Android applications (apps) developers increasingly rely on code obfuscation techniques to hinder reverse engineering and protect intellectual property. However, obfuscation also reduces the effectiveness of static analysis and vulnerability detection tools, creating challenges for Android security analysis. Existing approaches for detecting obfuscation in Android apps predominantly rely on handcrafted heuristics, engineered features, or task-specific learning pipelines, which may struggle to generalize across evolving obfuscation strategies. This paper presents a large-scale empirical study investigating the capability of Large Language Models (LLMs) to detect obfuscation in Android apps through semantic reasoning. Our study evaluates whether off-the-shelf LLMs can identify obfuscated code without relying on handcrafted rules, predefined signatures, or dedicated model training. The empirical evaluation is conducted on both a controlled benchmark containing an app obfuscated with multiple techniques and a real-world dataset of Android apps collected from Google Play. The study further examines the impact of prompt design, model selection, and decision thresholds across several open-weight and proprietary LLMs. Finally, the analysis compares LLM-based reasoning with existing SAST-based obfuscation-detection approaches and discusses the broader implications and limitations of applying LLMs to Android security analysis.

Deepfakes are artificially generated images, audio, or videos that threaten privacy, security, and information integrity. Detecting such content is crucial for countering disinformation, as the latest models generate highly realistic content. While spatial- or frequency-based approaches achieve good detection rates on Generative Adversarial Networks (GANs)-based generated deepfakes, they often struggle with recent diffusion model-generated images. In particular, existing approaches rarely exploit complementary multi-domain representations or systematically evaluate cross-generator robustness. To address these challenges, we propose a multi-domain deepfake detection framework called SGFF-Net (Spatial-Gradient-Frequency Fusion Network) that integrates spatial, gradient, and DWT (Discrete Wavelet Transform)-based frequency representations within a dual residual learning architecture. Experimental results show that the SGFF-Net achieves 98.95\% accuracy in intra-dataset evaluation and improves performance in both cross-model (70.46\%) and cross-paradigm (69.94\%) settings. Incorporating multi-source training and data augmentation further enhances robustness, increasing accuracy from 70.46\% to 79.80\% in cross-model evaluation, from 69\% to 78\% in cross-paradigm evaluation, and from 61.50\% to 75.80\% on real-world data. Unlike single-domain detectors, the SGFF-Net learns complementary forensic cues across spatial, gradient, and wavelet-frequency domains, resulting in greater robustness under cross-generator and cross-paradigm evaluation. The results further show that combining multi-domain representations with data diversity and augmentation substantially improves generalization, providing practical insights for developing more reliable deepfake detection systems.

In this paper, we introduce and evaluate Selective Field Transmission (SFT), a middleware mechanism that decouples transmission content from statically defined message types in publish-subscribe systems. Industrial and robotics developers often face a dilemma: They can follow established best practices and use standard message types, such as in the Robot Operating System 2 (ROS 2) and COVESA projects, to benefit from reusable and interoperable interfaces, or they can introduce proprietary, project-specific message types tailored to receiver requirements to reduce bandwidth. SFT resolves this trade-off by dynamically adapting the transmitted message components to each receivers actual needs while preserving unmodified standard interfaces. Receivers declare or automatically derive the required message components, which are communicated to the publisher. The publisher then serializes and transmits only the required component subset per receiver with minimal developer intervention. Our evaluation shows that SFT achieves significant bandwidth reductions without measurable per-message latency overhead, with savings proportional to the number and size of unused fields.

The rapid evolution of Time Series Foundation Models (TSFMs) has advanced zero-shot forecasting across diverse domains. Inspired by the current form of Large Language Models, future TSFMs may be offered as commercialized, closed-source API services. However, many existing online adaptation methods still rely on white-box access for parameter fine-tuning or gradient backpropagation. This paradigm mismatch raises a question: In black-box online adaptation for TSFMs, what should we learn? We answer this with an insight: the predictive errors of the base model are conditioned on both the input and output of the base model (i.e., the context of errors). To validate this insight, we propose ORCA (Online Residual Contextual Adaptation). We conduct extensive experiments across 5 state-of-the-art TSFMs and 8 datasets to demonstrate the effectiveness of our approach. Furthermore, through ablation studies, we quantitatively analyze the impact of different adapter learning hypotheses on the final adaptation performance in black-box online adaptation. Code available at https://github.com/Fifthky/ORCA.

Agentic AI can support unmanned aerial vehicle (UAV) autonomy by providing high-level recovery reasoning when local waypoint- or setpoint-based execution encounters blocked passages, repeated no-progress behavior, or mission-level ambiguity. On physical UAVs, however, remote reasoning is most useful when it is invoked selectively, since each call introduces latency, resource cost, backend uncertainty, and a need to validate the returned decision. This paper presents Persistent Mission Runtime (PMR), a UAV recovery framework that keeps the mission loop and safety-critical execution local while using an external agentic reasoner only as an on-demand recovery module. The reasoner selects from predefined recovery skills, and each returned decision is parsed, verified, safety-filtered, and mapped to local executor actions before it can affect flight. PMR introduces learned Cognitive Value of Invocation (learned-CVI), a compact admission gate that estimates when remote agentic reasoning is likely to improve near-term mission progress enough to justify its operational cost. Across a fixed 400-run Gazebo/PX4 benchmark with eight scenarios, learned-CVI raises hard/ambiguous-regime success from 5.0% under local-only autonomy to 95.0%, outperforms one-shot and periodic reasoning baselines by 20.0 and 32.5 percentage points, and reduces remote-agent calls by 16.7% and logged tokens by 29.2% relative to a manually tuned rule-based invocation baseline.

For robots to work safely in household environments, they need to be compliant and react to torque and force feedback during contact. However, the majority of existing data collection pipelines still lack the ability to capture force and torque data for learning active compliant policies. In this paper, we present Universal Manipulation Exoskeleton (UME), an upper-limb exoskeleton that provides real-time haptic torque feedback while recording whole-arm configurations and joint torque signals for teleoperation. With transparent torque feedback, human operators can even unsheathe kinematically constrained objects while blindfolded. UME is low-cost, lightweight, and portable. Equipped with an embedded IMU, it enables teleoperation for mobile manipulation. With our proposed universal retargeting algorithm, UME can teleoperate a range of robots, including the 7DoF OpenArm, 7DoF Franka, and 6DoF X-ARM. We demonstrate that this combination of capabilities enables learning bimanual, whole-body, and active compliant policies that operate effectively in highly constrained spaces. The learned robust autonomous policies achieve high success rates across a variety of tasks, including long-horizon mobile manipulation, force-mediated box flipping, visually occluded box pushing, and space-constrained tabletop manipulation. Videos, code, and additional information can be found at https://ume-exo.github.io.

Accurate prediction of protein-ligand binding affinity is essential for structure-based drug discovery. Recent geometric deep learning methods have achieved promising performance by representing protein-ligand complexes as three-dimensional graphs. However, most existing approaches mainly rely on static interaction geometry from a single bound conformation, while neglecting molecular flexibility and binding-induced conformational changes. To address this limitation, we propose a curvature-informed potential energy surface (CPES) graph neural network for protein-ligand binding affinity prediction, which incorporates physics-informed curvature representations to model conformational flexibility. CPES first derives curvature spectral descriptors from the Hessian of the potential energy surface evaluated at equilibrium configurations, whose eigenvalues define the local principal curvatures of the potential energy surface. It then uses spectral cross-attention to compare the unbound ligand and protein with the bound complex, thereby capturing binding-induced changes in conformational dynamics. In parallel, hierarchical protein-ligand interaction representations are learned from static structural features through geometry-aware message passing, soft clustering, and bidirectional cross-attention. Finally, CPES fuses the curvature-informed dynamic representations with static interaction representations for affinity regression. Extensive evaluations on multiple benchmark datasets demonstrate that CPES achieves improved predictive performance and offers physical interpretability.

The emergence of Large Language Models (LLMs) has inspired the vision of generating bespoke crystal materials directly from natural-language instructions, enabling users to design materials through intuitive, conversational interaction. Existing text-to-crystal generative models represent important early steps toward this goal, but they suffer from two critical limitations: (i) restricted input formats that require highly structured descriptions (e.g., chemical formulas), and (ii) one-directional generation, where models can map text to crystal but cannot perform the inverse. These limitations prevent fully conversational workflows and hinder alignment with users' inherently ambiguous and evolving desiderata. We address these challenges with LapidaryEngine, the first model to support fully conversational crystal generation. LapidaryEngine accepts free-form natural-language requests and performs iterative refinement and editing in a dialogue-like manner. The key innovation is a pivot representation, a third, intermediate form that enables bidirectional translation between text and crystal structures despite the absence of direct paired datasets. Leveraging this pivot allows robust interpretation of user feedback and precise structural control. We demonstrate LapidaryEngine across diverse tasks, including insulator discovery, stability optimization, compositional modification, and structural editing, showcasing its ability to align generated materials with user intent in an interactive manner.

LLMs are increasingly deployed as agents that interact with external environments and observe feedback such as execution results, error messages, and tool outputs. A well-functioning agent should be able to leverage this feedback to accurately assess its own performance. Yet we find a persistent reflection gap: LLM agents tend to mis-assess their own outputs after observing concrete environment feedback -- even for questions they correctly answered -- and standard RL barely helps due to a credit-assignment mismatch. To close this gap, we propose RefGRPO, a simple yet effective fix that augments standard RL algorithms with two key ingredients: a free calibration bonus computed by contrasting the agent's own reflection with the actual outcome (requiring no additional reward model, LLM judge, or external annotation), and a dynamic schedule on its coefficient. Compared to standard RL baselines, our method simultaneously improves reflection calibration (e.g., reduces underconfidence rate $44.4\% \to 7.7\%$) and task accuracy (e.g., $75.1\% \to 76.5\%$) on text-to-SQL across five benchmarks. The resulting calibrated reflection turns the agent into its own verifier grounded in environment feedback, which further enables (i) better self-improvement that uses reflections as pseudo-rewards without outcome supervision, and (ii) more effective test-time selective prediction by committing only to rollouts flagged as correct.

Large language models (LLMs) are increasingly deployed in privacy-sensitive domains, where users must balance the risk of data exposure through external APIs against the high computational cost of local deployment. Split learning has therefore emerged as a promising paradigm for LLM fine-tuning and inference under limited local resources. However, it introduces new privacy risks. Prior work primarily studies leakage of private input prompts, typically via inversion attacks on intermediate representations, while the potential for sensitive information leakage through generative response outputs remains largely unexplored. In this work, we unveil novel vulnerabilities of Split-LLM by presenting Patched Model Inversion with Dual-Sided Initialization (PIDI), a two-stage attack that simultaneously targets both private input prompts and output responses in Split-LLM settings. It combines dual-sided initialization with a patched inversion strategy to tackle long sequences, substantially outperforming prior inversion methods. To counter threats from both sides, we further propose the Adapter-based DualGuard with Mutual Information Defense (ADMI), which integrates an adapter-based local warmup strategy and mutual information regularization to provide a strong empirical privacy protection with minimal impact on task performance. Extensive experiments across diverse tasks and models demonstrate that ADMI effectively defends against PIDI and other state-of-the-art inversion attacks. Our code is publicly available at https://github.com/FLAIR-THU/VFLAIR-LLM.

In this paper, we evaluate the extent of undisclosed LLM-generated content in texts from the parliaments of the United Kingdom and Sweden. In many areas, such as in journalism or in academic writing, there are often requirements to clearly disclose whether AI tools, such as LLMs, have been used. In the case of parliamentary texts, the guidelines on disclosure of AI use are more vague. However, in order to maintain transparency and retain public trust, it is generally recommended that parliamentarians should state whether or not they have used AI when writing texts, such as parliamentary motions. Here, we train an interpretable (glass-box) text classifier using pre-LLM parliamentary texts and LLM-generated versions of such texts. We then apply the classifier to a test set containing recent parliamentary texts, finding a steady increase in undisclosed LLM use, in both parliaments, from 2022 onwards.

Large Language Models (LLMs) have advanced Automatic Heuristic Design (AHD) by enabling heuristic generation through reasoning and code synthesis. Existing LLM-based AHD architectures mainly follow two paradigms: Natural Evolution, which uses crossover and mutation to explore heuristic programs, and Metacognitive Evolution, which refines reasoning through reflection. However, Natural Evolution discards reasoning traces, weakening knowledge inheritance and exploitation, while Metacognitive Evolution lacks population-level recombination, limiting exploration and increasing the risk of premature convergence. These limitations reduce search efficiency, stability, and solution quality on complex problems. To address this gap, we propose MeEvo, a dual-layer AHD framework that cyclically couples Natural Evolution and Metacognitive Evolution. Natural Evolution explores heuristic code while recording reasoning traces, fitness values, and errors into a shared history; Metacognitive Evolution then reflects on this history to generate improved heuristics that re-enter the parent pool for the next cycle. This design enables population-driven exploration and reflection-driven refinement to reinforce each other. Experiments on five optimization problems with two LLM backbones show that MeEvo achieves stronger and more stable performance than existing LLM-based AHD architectures, especially on complex constrained tasks.

Open platforms increasingly route tasks among heterogeneous LLM agents--differing in base model, scaffold, and tool stack--whose competence varies sharply by skill: an agent excellent at one skill may be useless at another. The standard reputation approach summarizes each agent by a single global trust score, but that scalar is the wrong object here, because routing every task to the globally most-trusted agent leaves the value of specialization unclaimed. We study skill-conditional trust R(i | k)--the trust to place in agent i for a task requiring skill k, rather than one score per agent--and pose three falsifiable questions: when is conditioning worth it, how much cross-skill evidence should be borrowed, and whether that borrowing is safe. A controlled phase-diagram analysis answers the first two: conditional trust wins only in a specific regime--high agent heterogeneity, sparse per-skill evidence, and correlated skills--and the coupling strength beta that buys this data efficiency is dual-use, because the same cross-skill borrowing is also a laundering channel. On a public benchmark of 14 genuinely heterogeneous AppWorld agents, real pools land inside the beneficial regime--a small but genuine gain, with the per-skill best agent genuinely changing across skills. We then show that an attacker with cheap evidence in one skill and none in a target skill hijacks the conditional router, driving routing regret from 0 to 0.94 on a pool our zero-cost Conditional Information Value Test (CIVT) rates GREEN--while the ungated trust verdict it contaminates reads -0.06 instead of the honest +0.19. A zero-evidence gate bounds the attack but does not eliminate it; we characterize the residual cost under an explicit budget. We do not claim Sybil-resistance--we quantify the trade-off.

Large language models are increasingly deployed as human simulators for interactive evaluation and social simulation. Yet helpfulness-driven post-training pulls them toward a homogeneous, overly agreeable assistant register, creating a behavioral Sim2Real gap. We present OdysSim, the largest open systematic investigation of behavioral foundation models, i.e., models trained to simulate human behavior at scale. We propose SOUL, a taxonomy of five capability axes (CONV, SS, COG, ROLE, EVAL) that unifies 62 datasets and 23 benchmark tasks under one framework. Specifically, we curate the OdysSim corpus (21.4M interactions, 10B tokens, retrofitted with back-generated social contexts), construct the SOUL-Index benchmark, and develop an end-to-end training recipe combining midtraining, task-specific RL, and expert distillation. The resulting open 8B OSim model ranks first or tied-first on 8 of 23 tasks, outperforming any individual frontier model by this count, with the strongest gains on conversational and social tasks. Its outputs are also more human-like in length, formatting, and word choice, and it transfers zero-shot to out-of-distribution user simulation on $τ$-bench, nearly matching real users on reaction alignment (93.2 vs. 93.5). We further show that LLM-as-judge RL induces reward-hacking patterns, and that our detectors can mitigate them during post-training. Together, our findings suggest that behavioral foundation models require rethinking the LLM training paradigm. We release all artifacts to support future research.

Kalman filters based on the Embedded Latent Transfer Operators (ELTO) emerge as novel statistical tools for sequential state estimation. However, a critical limitation stems from their use of simplified noise models, which fail to dynamically adapt to non-stationary processes. To address this limitation, we introduce an ELTO-based Bayesian filtering approach with a new structured parameterization for the filter's noise model. This parameterization enables structured noise adaptation, which couples the data-driven learning of an optimal time-invariant noise model with dynamic parameter adaptation that responds to changes in dynamics within non-stationary processes. Empirical results show that our structured noise adaptation improves the filter's dynamic state estimation performance in noisy, time-varying environments.

This paper presents a two-stage Hybrid Classical-Quantum (HCQ) pipeline for binary Alzheimer's disease (AD) classification from 3D T1-weighted structural MRI volumes, where the classical and quantum components are designed to complement each other rather than operate independently. A supervised 3D $β$-variational autoencoder (VAE) is trained end-to-end under voxel-wise reconstruction, KL-divergence, and focal classification losses that compress each 3D MRI volume (resized from 152 x 184 x 152 to 96 x 96 x 96) into a 64-dimensional latent code. Partial Least Squares (PLS) regression selects the six components in the latent code that best separate Alzheimer's Disease (AD) from cognitively normal (CN) subjects and rescales them into rotation angles, which are encoded onto a six-qubit register using the ZZ quantum feature map to give us the respective quantum states. The input to a precomputed-kernel Support Vector Machine (SVM) is an N x N Gram matrix (N = 308), created by calculating the overlap between every pair of quantum states. The novelty of this work lies in the fact that the quantum kernel operates directly on disease-aware features that are learned end-to-end by a supervised autoencoder, rather than on pre-extracted inputs. On 308 ADNI-1 subjects, consisting of 137 AD and 171 CN subjects, the baseline achieved 67.2% accuracy and 0.759 AUC, while the stability-enhanced variant reached 72.1% accuracy and 0.799 AUC with cross-fold variance halved. 3D Grad-CAM further helped validate our model's focus on brain regions linked to Alzheimer's. The HCQ pipeline could serve as a general-purpose framework for diagnostic classification across biomedical imaging domains that present similar challenges for classical approaches.

Auto-bidding is a core component of real-time advertising systems, where decisions must optimize long-term performance under budget and cost constraints, while online exploration is prohibitively risky. Offline reinforcement learning and, more recently, Transformer-based sequence modeling have shown promise for learning bidding policies from logged data, but their unimodal and purely parametric formulations often collapse multiple effective bidding strategies into suboptimal averaged actions and perform unreliably under sparse or long-tail traffic. To mitigate these limitations, we propose DRIVE (Distributional and Retrieval-Augmented Bidding with Value Evaluation), a unified Transformer-based framework that decouples candidate action generation from decision making for offline auto-bidding. DRIVE combines distributional action modeling, retrieval-augmented candidate generation from high-quality historical decisions, and value-based evaluation to select the most promising bid at inference time. Extensive experiments on AuctionNet and additional offline reinforcement learning benchmarks demonstrate that DRIVE consistently improves bidding performance and generalizes well across multiple Transformer-based methods.

We present CORD-SLS, a real-time control method for safe deformable object manipulation, with a focus on ropes and cloth. At its core is a GPU-parallel differentiable simulator with contact smoothing which enables efficient gradient-based planning through intermittent contact. To robustly satisfy constraints under model and sensing uncertainty, we develop a real-time, GPU-parallel output-feedback robust model predictive control (MPC) algorithm that plans with this simulator. We further show that the simulator accelerates model-based RL for training neural manipulation policies. To improve real-world robustness, we use conformal prediction to calibrate visual-feedback and perception-error bounds for MPC, producing reachable tubes that enable high-probability safe control. We evaluate CORD-SLS on high-dimensional, contact-rich rope and cloth manipulation tasks in simulation and hardware, including obstacle avoidance, routing, folding, and smoothing. Across settings, CORD-SLS achieves millisecond-speed planning, exceeding baselines in safety, speed, and task success.

Large-scale neural network training increasingly relies on matrix-aware optimizers that exploit the structure of weight parameters beyond element-wise adaptation. However, existing matrix-aware methods such as Muon have an underappreciated vulnerability: their core operation, Newton-Schulz iteration, depends critically on input conditioning, yet the raw momentum matrices exhibit severe coordinate-wise scale heterogeneity. In this paper, we first verify this scale heterogeneity through a chi-square uniformity test, showing that intra-matrix scale imbalance is prevalent across Transformer layers and that coordinate whitening effectively corrects it. Motivated by this finding, we propose Zeta, a dual whitening optimizer that applies coordinate whitening and spectral whitening in a strictly ordered pipeline. The ordering is not a tunable choice but follows from a mathematical dependency: coordinate whitening establishes the statistical isotropy that spectral whitening requires to function reliably. We further prove that this dual pipeline strictly reduces orthogonalization error relative to pure spectral methods by improving the condition number of the input. Empirically, Zeta matches or surpasses strong baselines across language modeling (0.6B to 8B parameters), mixture-of-experts architectures, and vision tasks, demonstrating that resolving scale imbalance before orthogonalization leads to faster convergence and better generalization. Code is available at https://gitcode.com/kevin259/MindSpeed.

Physics-informed neural networks (PINNs) are meshless and carry moving geometry and topology change through resampling of collocation points; the finite-element method (FEM) is the workhorse for boundary-fitted discretisations. Coupling the two across a shared interface promises the best of both, yet existing PINN-FEM schemes are validated only empirically. We put the coupling on a domain-decomposition footing: viewing each solver as a Steklov-Poincaré (trace-to-flux) operator, we transfer the classical Dirichlet-Neumann (DN) divergence diagnosis and its Robin-Neumann (RN) cure, including a closed-form, sweep-free interface impedance, and prove a PINN-specific contraction theorem: a trained network realises only a perturbed Steklov operator with a per-step training residual, and RN still contracts, with no shared-eigenbasis hypothesis, to a floor set by the achieved training loss. Because a PINN has no stiffness matrix, we introduce a Fourier-mode interface probe that recovers the network's resolvable Steklov eigenvalues to within 0.5% and doubles as a diagnostic of the network's spectral cap. The theory predicts measured PINN-FEM contraction rates to within 7% on 1D and 2D Poisson couplings, and a two-slab analogue of the large-added-mass regime shows RN's per-mode impedance matching winning decisively where tuned scalar relaxation saturates. We demonstrate the framework on a Stokes/rigid-disc problem with Alart-Curnier contact: the meshless PINN fluid absorbs the topology change at contact by collocation exclusion alone, no remeshing and no cut cells, and the static-equilibrium contact reaction matches the submerged weight to 0.4% under mesh refinement. We quantify remaining limitations: the warm-started PINN drifts off the Stokes manifold over long horizons, and matched FEM-FEM benchmarks attribute pre-impact squeeze-film signatures to PINN under-resolution.

We present CacheRL, a system for training small agent foundation models that achieves 92 percent process accuracy on multi-step tool-calling tasks, approaching GPT-5's 94 percent while requiring 100 times less compute. Our approach addresses three challenges in practical agent training: transferring tool-calling knowledge from large models at scale, enabling reinforcement learning without costly live tool execution, and learning robustly from noisy cached environments. CacheRL introduces three key innovations. First, a hybrid thinking trajectory pipeline augments agent trajectories with LLM-generated reasoning traces, producing training examples that teach models not only what tools to call but also why. Second, the CacheAgentLoop eliminates live execution costs through a three-tier fuzzy cache while preserving trajectory fidelity using token-level masking. Third, a cache-tier-aware reward dynamically adjusts answer-quality weights to avoid penalizing models for cache-induced limitations. Through iterative supervised fine-tuning (SFT) and Group Relative Policy Optimization (GRPO), CacheRL improves Qwen3-4B-Thinking's validation reward from 0.43 to 0.78. On public agentic tool-calling benchmarks, our model achieves competitive performance against frontier models such as GPT-5. Ablation studies show that removing knowledge transfer reduces performance by 41 percent, while cache-aware rewards contribute a 17 percent improvement. Interestingly, reinforcement learning improves training stability but yields limited gains beyond strong supervised fine-tuning, suggesting that data quality and reward design play a more important role than complex optimization methods in building practical small agent models.

Geometry problem generation is useful for AI-assisted education and multimodal mathematical reasoning, but reliable synthesis remains difficult because the problem statement, diagram, constraints, and solution should be mutually consistent. Existing methods often trade off controllability and reliability: seed-based rewriting is flexible but weakly verifiable, whereas diagram-first construction improves validity but is less suited to arbitrary user-specified constraints. We introduce VeriGeo, a controllable geometry generation framework grounded in executable reasoning traces. Given user constraints such as target concepts and difficulty, an Author agent generates a problem and diagram, and a Solver agent produces a proof-aligned solution. Both agents use a shared action sequence that connects natural language, diagrams, geometric constraints, and proof steps into a verifiable representation. A three-stage pipeline checks numerical consistency, analytical realizability, and global consistency, using verification-guided reflection to repair recoverable failures and reject unrecoverable ones. Across five LLM backbones, raw generations frequently fail these checks, while VeriGeo repairs a substantial fraction of the invalid attempts. Supervised fine-tuning on 8.7k examples generated by VeriGeo achieves the best reported GeoQA performance among end-to-end multimodal LLM-based solvers, and obtains strong results on PGPS9K and MathVista-GPS, demonstrating the effectiveness of verified synthetic data for improving multimodal geometry reasoning.

Multimodal Attributed Graphs (MAGs) model real-world entities by coupling graph topology with heterogeneous attributes such as text and images. They support graph-centric tasks requiring structural and class-discriminative representations, and modality-centric tasks requiring fine-grained cross-modal correspondence. However, existing MAG methods often rely on fixed graph contexts or uniformly fused representations, causing task-agnostic propagation and over-compressed fusion that hinder diverse task requirements and modality-specific evidence preservation. To address this, we propose CoMAG, a unified MAG backbone that learns task-adaptive reliable contexts and modality-preserving alignment within them. CoMAG first conducts Reliable Context Learning by estimating edge reliability from multimodal semantic consistency, complementing raw topology with semantic neighbors, and selecting context components through a task-aware gate. It then performs Modality-preserving Hop-token Alignment by maintaining modality-specific multi-hop trajectories, matching modality-hop tokens across modalities, and decoupling shared and private representations. Thus, CoMAG produces graph and modality representations from one forward pass while retaining modality-specific cues. We further analyze stable propagation, over-smoothing mitigation, and modality-collapse control. Experiments on nine OpenMAG datasets compare CoMAG with feature-only, graph-only, multimodal, and unified MAG baselines across graph-level prediction, modality matching, and graph-conditioned generation. Results show that CoMAG achieves the best reported performance, demonstrating that task-adaptive reliable contexts and modality-preserving alignment improve structural prediction, cross-modal matching, and graph-conditioned generation while retaining sparse edge-linear complexity.

Raman spectroscopy provides label-free, chemically specific characterization of biological systems and has become an important tool for cancer diagnosis, molecular subtyping, microbiological identification, and intraoperative decision support. Biomedical Raman spectra are, however, high-dimensional, noisy, and affected by fluorescence background, acquisition variability, and biological heterogeneity, making robust computational analysis essential. This review examines the role of machine learning across the biomedical Raman spectroscopy pipeline, from preprocessing and signal correction to unsupervised structure discovery, supervised diagnosis and molecular stratification, representation and transfer learning, explainability, biomarker discovery, and multimodal integration with imaging, pathology, and molecular profiling. Emphasis is placed on the use of machine learning not only for diagnostic classification, but also for biologically interpretable and clinically actionable analysis. We also discuss the main barriers to clinical translation, including limited dataset sizes, inter-instrument variability, inconsistent preprocessing, insufficient external validation, reproducibility concerns, and limited sharing of software, data, and metadata. We argue that progress will require methodological advances together with standardization, robust validation, explainability, and deployment-ready analytical frameworks. By integrating methodological, biomedical, and translational perspectives, this review outlines key directions for developing reliable and clinically deployable Raman-AI systems.

Fuzz drivers are essential components of greybox fuzzing, as they encapsulate target interfaces, define test spaces, and largely determine fuzzing effectiveness. Existing fuzz drivers typically rely on crash-based oracles for security testing, overlooking library functionality and limiting bug detection capability. In this paper, we present the first study on metamorphic-based fuzz oracle enhancement (MFOE), which augments existing fuzz drivers with metamorphic-based oracles derived from metamorphic relations (MRs). Since constructing and integrating such oracles requires substantial domain knowledge, automating MFOE is challenging. To address this challenge, we propose MetaFOE, an LLM-based framework that automatically generates and integrates metamorphic-based oracles. We evaluate MetaFOE on OSS-Fuzz drivers using three modern LLMs and five prompt strategies. MetaFOE generates 3,475 MRs, of which 77.3% are applicable, and implements 12,351 meta drivers, with 6,228 being valid. After three hours of fuzzing, the valid meta drivers improve edge coverage by an average of 18.7% and trigger 1,528 unique crashes. Our results demonstrate both the effectiveness of metamorphic-based oracle enhancement and the feasibility of using LLMs to automate MFOE, providing valuable insights for advancing greybox fuzzing.

Protein-ligand binding affinity (PLA) prediction is critical in drug discovery. Despite the notable advancements in machine learning-based approaches, existing methods struggle to jointly characterize local geometric organization and globally coordinated cross-molecular interactions, limiting their ability to model complex binding mechanisms. Here, we propose RicciBind, a geometric representation framework that integrates curvature-guided hierarchical structure learning with optimal transport (OT)-based cross-domain alignment to model molecular interactions. Specifically, RicciBind leverages Ricci curvature to capture local interaction tightness within molecular structures, enhancing structural awareness and organizing atomic interactions into curvature-aware hierarchical representations. An OT-based cluster matching mechanism then aligns protein and ligand clusters across heterogeneous domains under geometric constraints, enabling globally consistent correspondences and revealing higher-order interaction patterns beyond local neighborhoods. By coupling curvature-guided structure encoding with OT-driven cross-domain alignment, RicciBind effectively models complex interaction semantics and substantially improves both the accuracy and interpretability of binding affinity prediction. Extensive experiments demonstrate that RicciBind achieved superior predictive performance and generalization across PLA benchmarks and virtual screening tasks. Ablation studies further confirmed the essential role of Ricci curvature in enhancing molecular interaction representations.

Cities deliver basic services through mixed public-private facility networks, including schools, clinics, transit providers, and subsidized service points. In these systems, planners often observe where households go, but not the latent cost function through which they trade off factors such as distance, price, and institutional access. We study this urban problem through school choice in the Philippines, where the country's largest national education subsidy is intended to redirect learners from congested public schools to participating private schools. Treating school-to-school enrollment flows as an entropic optimal transport plan, we recover latent choice costs using two complementary inverse optimal transport models: an interpretable distance-banded model with a subsidy term, and a neural cost model trained through a differentiable Sinkhorn forward pass. Applied to 283{,}016 learner trips across 23{,}820 observed flows in the most populated region, the framework estimates a subsidy-equivalent distance, $λ^{(k)}$, interpreted as the kilometers of perceived travel cost offset by the subsidy. The case demonstrates how administrative origin-destination data can be transformed into interpretable planning metrics for accessibility-aware subsidy design, facility siting, and urban service allocation.

Learning systems based on IF-THEN rule representations readily offer interpretability, making them a crucial focus in contemporary AI research. A key objective for such rule sets is to achieve both high discriminative power and interpretability. While existing state-of-the-art algorithms implicitly prioritize predictive accuracy, they often fall short on one or more quality metrics that ensure interpretability, such as coverage and parsimony of rule sets. Motivated by this, this paper propose the development of CDPR, which aims to create highly accurate and interpretable rule sets for classification problems. To the best of our knowledge, this represents the first attempt to establish such an approach. In this study, we introduce two algorithms rooted in submodular maximization, which not only provide provable guarantees on coverage but also yield rule sets that are both discriminative and parsimonious. We empirically demonstrate that rule sets learned through our approaches achieve higher accuracy and interpretability and has more than a 2.5-fold improvement in average coverage rates when compared to the next best algorithm.

Context adaptation automates prompt engineering in LLM-based systems by iteratively revising tunable prompts from task feedback, without modifying model weights. Extending this paradigm to multi-LLM agentic systems is crucial: existing methods suffer from inaccurate credit assignment and lack convergence guarantees. We propose \textbf{G}raph-based \textbf{T}arget \textbf{B}ack-\textbf{P}ropagation (GTBP), a context adaptation framework for agentic workflows modeled as directed acyclic graphs. GTBP propagates local target outputs backward through the workflow graph and uses target--output discrepancies to guide a stage-wise prompt update mechanism. Theoretically, we show that GTBP's stage-wise prompt updates become stable over iterations, and that a sufficiently capable LLM optimizer can decrease the overall objective. Empirically, GTBP consistently outperforms strong baselines across three benchmarks while maintaining comparable computational cost.

Pruning promises a shortcut to strong small language models. In this work, we examine this promise by pruning Llama-3.1-8B at pruning ratios of 0.5--0.8 with six methods spanning depth, width, and sparse granularities, under two controlled token-matched settings. (1) With the same training token budget, pruned initialization consistently outperforms random initialization. This shows that the parent model provides a strong starting point, although the advantage narrows as the training token budget grows and as the pruning ratio rises, nearly vanishing at the highest pruning ratio we study. (2) When training from scratch is instead given the full token budget consumed by the whole pipeline, pruning at finer granularities still retains an advantage, while coarser structured pruning can be matched or surpassed. This suggests that the parent model transfers knowledge that additional training tokens alone cannot fully recover, but only at fine granularity. Taken together, our results yield a clear recommendation: with a large pretrained model in hand and a limited training token budget, pruning is better than training from scratch; when the training budget is not limited, training from scratch can be competitive for coarser pruning, so a large pretrained parent is not always necessary.

Large Language Models (LLMs) are increasingly deployed in healthcare settings, yet their tendency to hallucinate poses risks when clinical decisions are involved. This study examine whether LLMs recommend recently banned or withdrawn pharmaceuticals when answering clinical questions and tests an agent-based method for reducing such errors. We developed a five-agent "Trust but Verify" system using a single LLM backbone. To measure regulatory knowledge obsolescence, we created an adversarial dataset of 103 clinical MCQs where historically correct answers now refer to banned substances. This scale ensures statistical significance across various therapeutic classes. We evaluated three open-access model families (GPT-OSS, Llama-3, Falcon-3) under vanilla and agentic conditions. Performance was measured via pointwise score, label accuracy, Hallucination Error Rate (HER), and Component Fidelity (CF) score. We also observed clinical safety regression in proprietary models. In default configurations, all models showed high hallucination rates, consistently selecting banned drugs that matched training data patterns. Our proposed agentic architecture reduced HER by approximately 53% across models. Pointwise scores shifted from -0.25 (unsafe recommendation) toward 0.0 (appropriate refusal). The safety audit intercepted dangerous outputs even when models' parametric knowledge favored the banned substance. The proposed multi-agent framework offers a model-agnostic method for enforcing regulatory compliance that prioritizes patient safety over fluent text generation. Our work demonstrates a practical approach for deploying autonomous AI systems in safety-critical healthcare settings. It shows how real-time regulatory data can be integrated into LLM pipelines to support clinical decision-making.

Learning in biological multilayer neuronal networks offers insights that extend beyond the classical weighted-sum neuron model commonly used in artificial neural networks. This article presents an accessible guide to a mechanistic neuronal network model that more accurately captures aspects of biological computation while enabling a simple yet powerful mechanism for learning in multilayer neural networks. The proposed approach supports efficient online streamed learning and provides a practical alternative to backpropagation. We demonstrate its potential in an image classification task, achieving competitive classification performance. The approach's simplicity, biological grounding, and broad applicability highlight a promising path toward algorithms that unify mechanistic neuron models and machine learning.

Healthcare is essential, expert, and episodic by design - built around the roughly one hour per year a person spends with a clinician. The 8,759 hours outside clinical settings, where eating, sleeping, movement, medication, and stress actually shape long-term health, have no comparable infrastructure. The bottleneck for personalized health is not raw data or reasoning capability; it is the absence of that infrastructure layer. This paper introduces the Personal Care Utility (PCU): a layered, event-driven architecture proposed as the missing utility for everyday health, in the way that payments, networks, and power are utilities for their domains. PCU organizes continuous personal signals into semantically meaningful life events through a Personicle, estimates dynamic health state against personal baselines, reasons about cause and context, and routes guidance through an orchestrator that separates clinical decision logic, behavioral strategy selection, and natural-language expression. This separation lets large language models support reasoning and communication while keeping safety-critical clinical decisions grounded in validated evidence. We instantiate PCU for Type 2 Diabetes - turning CGM, meal, activity, medication, sleep, stress, and clinical data into glycemic events, individualized state estimates, causal explanations, and knowledge-grounded interventions. A day-in-the-life scenario shows the same infrastructure producing real-time nudges, weekly summaries, medication check-ins, silence, or deterministic safety alerts depending on context and risk. We close with how PCU generalizes to other chronic conditions and the governance questions any always-on personal health utility must address. The result is a blueprint that treats personalization not as a final messaging layer, but as an architectural property of everyday health guidance.

Large Language Models (LLMs) are increasingly adopted as backbones for Generative Recommendation (GR), promising access to pretrained world knowledge. Yet reliably invoking this knowledge for GR remains poorly understood. A key obstacle is that LLM-based GR typically represents items with Semantic IDs (SIDs), disrupting LLMs' natural-language reasoning interface because these tokens are unseen by the LLM during pretraining. Existing approaches address this with expensive multi-stage pipelines that ground SIDs and elicit explicit rationales, but offer limited insight into when and why each stage is necessary. In this work, we systematically decompose explicit reasoning training pipelines for LLM-based GR, revealing three key limitations: weakened world-knowledge verbalization, misalignment between SID and natural-language token embedding spaces, and sensitivity to rationale quality, all of which hurt explicit reasoning performance. To circumvent these issues, we propose PauseRec, a lightweight implicit reasoning paradigm tailored for GR. PauseRec is exceptionally practical, avoiding costly reasoning trace acquisition and reasoning alignment training, leading to a multitude of benefits: (1) it outperforms standard explicit CoT methods by up to 6.22%, (2) it reduces training cost by up to 65% GPU hours, and (3) it speeds up inference by up to 71.3%. These results position PauseRec as a lightweight alternative to explicit rationale generation, enabling more effective and efficient LLM-based GR.

Sound events are entities with semantic identities, locations, and trajectories, but current audio-language models usually reason about clips as global event content. Conversely, sound event localization models track source directions over time but offer limited semantic coverage for language reasoning. To address this gap, we introduce ST-AudioQA, a spatio-temporal audio QA dataset and benchmark built from first-order ambisonic (FOA) renderings of static and moving sound sources. Each scene provides source identity, activity, direction, distance, and motion metadata, enabling dense trajectory supervision and questions about what is sounding, where it is, how it moves, and how sources relate. We further propose ST-Audio Encoder, a time-resolved FOA audio encoder that learns event semantics together with source trajectories, and ST-AudioLM, which connects the audio tokens from the encoder to an LLM for spatio-temporal audio QA. Experiments show that this representation improves the semantic-localization tradeoff and yields stronger reasoning performance than static spatial and localization-oriented baselines.

Full waveform inversion (FWI) recovers subsurface velocity from seismic recordings by solving a severely ill-posed, nonconvex PDE-constrained optimization. Classical regularizers stabilize the inversion but fail to reproduce realistic geological structures; recent diffusion-prior methods improve realism at the cost of a fragile trade-off between data fidelity and prior consistency. We propose Decoupled Latent Optimization (DLO), which relaxes the standard latent-optimization formulation into a quadratic-penalty objective over an auxiliary physical variable and a latent variable. The data-fidelity gradient acts in physical space, the diffusion sampler contributes only through a decoded prior sample, and the standard smoothed-velocity initialization of classical FWI is preserved. On the OpenFWI benchmark, DLO outperforms classical regularizers and existing diffusion-based methods under clean, noisy, and missing-trace acquisitions. The prior, trained on 70*70 OpenFWI models, transfers directly to the Marmousi and Overthrust benchmarks, where DLO recovers intricate fault structures and remains robust to initialization smoothing and measurement noise.

Safe coordination problems surface in multi-agent reinforcement learning when global safety cannot be enforced by any agent unilaterally: the admissibility of one agent's action may depend on the dynamics of other agents. Decentralised shields can enforce safety at runtime, but purely factorised permissions often exclude optimal team behaviour that is safe only through coordination. We study deterministic safety guarantees for agents trained and deployed under decentralised execution, recovering team-optimal safe behaviour without centralised runtime control. Agents have a shared global specification $φ$ in the safety fragment of Linear Temporal Logic ($\mathsf{LTL}_{\mathsf{safe}}$ ), and select among tuples of local $\mathsf{LTL}_{\mathsf{safe}}$ obligations whose conjunction implies the global specification $φ$. Each agent may rely on the other agents' local obligations as assumptions because the whole contract tuple is certified simultaneously and allows projection into local action masks. At learning time, a non-stationary multi-armed bandit chooses among a library of local $\mathsf{LTL}_{\mathsf{safe}}$ obligations to select the tuple that optimises team reward, all without forgoing end-to-end safety. We evaluate the approach across 6 environments and 15 algorithmic variants.

LLM-based query rewriters in production face a tension: the training reward must reflect how the rewrite is consumed by the production ranker, yet the training procedure must be cheap enough to support continuous redeployment as data drifts. We present CoRe (Context Relevance), such a system, redeployed weekly for over five months in a major short-video search engine. Our reward uses the deployed multimodal relevance model as its source and a multiplicative ratio form mirroring the production fusion algebra, closing the simulation-production gap that offline reward proxies leave open. A semi-online Mixed Preference Optimization loop makes this reward affordable at multi-million-instance weekly scale: a DPO-style pairwise objective restricts the gradient pass to a small top-k/bottom-k subset of sampled trajectories, and a phase structure reduces trainer/inference-server parameter syncs from per-step to per-phase. An automated promotion gate over reward-like and stability metrics detected and recovered from a real reward-hacking incident in production. Rewriter output is consumed as parallel relevance signals at recall, rawrank, and finerank without displacing the original signals, bounding rewriter-failure blast radius. Online A/B from two sequential production launches, first deploying the rewriter at finerank, then extending consumption to recall and rawrank, delivers statistically significant reductions in change-query rate on rewrite-impacted queries, with all headline relevance and engagement metrics moving in the expected direction.

Inversion-based image editing offers flexible and training-free control but still struggles with inversion accuracy and the trade-off between editing fidelity and background preservation. While recent methods improve inversion formulations or attention interactions, the role of textual conditioning in shaping diffusion dynamics and editing behavior remains underexplored. We show both empirically and theoretically that the precision of textual conditioning influences inversion stability by modulating the geometry of the diffusion velocity field, while also affecting the consistency of cross-branch attention during editing. These effects directly impact background preservation and semantic fidelity. Building on this analysis, we propose SimEdit, a conditioning-aware framework with two complementary components: (a) conditioning refinement, which constructs conditioning signals with improved semantic precision and structural alignment to facilitate stable inversion and consistent attention manipulation, and (b) token-wise cross-branch attention control, which separates edit-relevant and structure-preserving components and modulates them asymmetrically during attention manipulation. Extensive experiments on PIE-Bench demonstrate that SimEdit consistently improves both inversion reconstruction quality and editing performance over previous attention-manipulation approaches. Our code is available at https://github.com/zju-pi/SimEdit.

Deployed knowledge-tracing models are typically frozen after training, yet systematic per-item logit bias arises, from limited per-item expressivity in backbone architectures and from post-deployment shifts in item properties, degrading prediction quality. Global post-hoc calibrators such as Platt scaling, temperature scaling, and isotonic regression improve probability estimates but leave discriminative ability, as measured by AUC, unchanged. This AUC invariance is a structural consequence of monotone score-only transforms; recovering the stranded discrimination requires conditioning on item identity. We propose SLC (State-space Logit Correction), which converts binary observations to Gaussian pseudo-observations via Laplace/IRLS, applies empirical-Bayes shrinkage through a Kalman smoother, and fits an offset-Platt link. The state-space formulation also yields a detectability bound that characterizes the Bernoulli information floor, explaining why temporal tracking provides no benefit at current data densities. Across four datasets, five backbones, and three seeds, SLC improves AUC on all four datasets and NLL on three, with the advantage concentrating on sparse items. Cross-domain controls suggest that the same phenomenon can arise beyond education when the deployed backbone leaves entity-level bias.

Byte-level tokenization enables language models to handle any Unicode input, but models can generate invalid UTF-8 sequences when encountering rare or unseen characters. We investigate the relationship between training scale and UTF-8 generation reliability with a 355M parameter model trained on 80B tokens from a balanced multilingual corpus of English, Japanese, Korean, and Chinese. We introduce multiple evaluation protocols that isolate UTF-8 structural validity from language modeling. UTF-8 validity convergence lags perplexity by a roughly a factor of two: perplexity stabilizes after 2.1B tokens, but UTF-8 validity requires 4.2B tokens. In context-free generation, rare characters achieve higher structural validity than common characters, suggesting over-specialization of frequent character representations. Through experiments, we observed that reliable UTF-8 generation is a distinct capability requiring evaluation beyond perplexity.

Auditory attention decoding (AAD) aims to infer the attended speaker from neural responses in multi-speaker acoustic environments and is a key problem for neuro-steered hearing systems. Although recent studies have achieved encouraging progress, existing AAD models still do not fully exploit frequency domain electroencephalography (EEG) information. In particular, most approaches introduce multi-band information through handcrafted feature extraction or direct cross-band feature concatenation, which mainly exploit frequency information at a shallow level and may overlook band-specific patterns and cross-band interactions. To address these limitations, this paper proposes FAConformer, a frequency-aware CNN-Transformer framework for AAD that explicitly integrates band-specific encoding and adaptive cross-band interaction. Specifically, FAConformer first decomposes EEG signals into multiple frequency bands and assigns each band to an independent CNN-Transformer encoder for band-specific modeling. The resulting band-wise features are then adaptively fused by a carefully designed frequency-aware attention (FAA) module that models cross-band dependencies by treating band-wise features as tokens. Further, band-wise auxiliary supervision (BAS) is introduced to prevent weakly contributing branches from being under-optimized during joint training. In this way, FAConformer performs frequency-aware modeling that more effectively exploits frequency domain information. Extensive experiments on two public AAD datasets with three decision-window lengths demonstrated that FAConformer consistently outperformed 12 competitive baselines, surpassing the current state-of-the-art model by 4.9%. Further analyses of band importance, ablation, and parameter sensitivity verify the effectiveness, robustness, and interpretability of the proposed framework. Code is available at https://github.com/wzwvv/FAConformer.

Fault diagnostics and recovery in smart factories is challenging because critical information is dispersed across manuals of multiple machines which are interconnected through the manufacturing process. Large Language Models (LLMs) can provide a promising approach. In this paper, we propose FactoryLLM, a safe and open-source AI playground designed for evaluating different LLM-based retrieval-augmented generation (RAG) models by analysing documents from multiple machines across the manufacturing process. FactoryLLM enables the user to configure the LLM, and assess performance when reasoning over multiple documents, through a dual evaluation setup using both RAGAS and NVIDIA's LLM-as-a-Judge metrics. FactoryLLM is safe because it allows users to run local or open-source LLMs without sharing sensitive industrial data, providing a controlled environment for experimentation. We demonstrate the efficacy of FactoryLLM through a case study which involves an Autonomous Intelligent Vehicle and its Mobile Planner software, evaluating three LLMs across 30 maintenance queries derived from approximately 600 pages of cross-machine documentation. The results suggest that FactoryLLM is effective in cross-machine document reasoning: every model achieved a groundedness score above 0.88. The full code and documentation for community to test FactoryLLM with their manufacturing specific scenarios are publicly available.

Large language models (LLMs) are increasingly evaluated for bias using adaptations of human psychological paradigms, yet methodological limitations-particularly the conflation of refusal behavior with task performance-have hindered clear interpretation. Here, we adapt the Implicit Association Test (IAT) to a controlled, forced-choice framework and introduce a two-stage modeling approach that separates response compliance from task-consistent classification. Across three contemporary LLMs (Claude Sonnet-4, Gemini 2.5 Pro, and GPT-5), we evaluate associative interference, defined as reduced task-consistency in incongruent relative to congruent conditions. While compliance with the structured response format was uniformly high, interference effects varied substantially across models and domains. Claude Sonnet-4 exhibited strong interference in the Gender--Career domain (DeltaP = 0.086, 95% CrI [0.026, 0.173]) and smaller but credible effects in Gender--Science. Gemini 2.5 Pro showed attenuated interference, and GPT-5 exhibited minimal or no detectable interference across domains. These findings demonstrate that IAT-style associative asymmetries are not a universal property of LLMs, but instead depend on model-specific characteristics. By isolating interference from compliance and modeling item-level variability, this study provides a principled framework for evaluating structured response patterns in LLMs. The results highlight the importance of model-specific assessment and suggest that associative interference can be substantially mitigated in modern systems.

The Differential Time-Varying Effect Model (DTVEM) of Jacobson et al. (2019) is a popular tool for finding the best time lag in intensive longitudinal data, but it assumes everyone shares the same lag structure. The original authors named fixing this as future work, and it clashes with the premise of modern clinical research, which is that people differ. We present DTVEM-RE, an extension that lets each person have their own lag coefficients, with two versions of the confirmatory step: a discrete-time hierarchical Bayesian VAR in Stan, which pools across people and gives calibrated uncertainty, and a continuous-time per-person Ornstein-Uhlenbeck model in ctsem, which handles unevenly spaced beeps directly. We report four results. A simulation shows the Bayesian version recovers the between-person spread tau_a with bias below 0.01 and coverage of 90 to 93 percent. On the Fisher et al. (2017) EMA dataset (N=40), person-specific lag-1 effects vary by an order of magnitude across three mood items, the Bayesian and GAMM estimates agree closely (r=0.87 to 0.92), and DTVEM-RE gives the best one-step-ahead prediction among four discrete-time methods. A multi-lag version shows all nine tau_k values have credible intervals excluding zero, and the lag where people differ most changes across items, something lag-1-only methods like mlVAR cannot detect. Finally, the two versions agree almost exactly on person-specific lag-1 estimates (r >= 0.995), differing only as shrinkage predicts. DTVEM-RE is, to our knowledge, the first person-specific implementation of DTVEM-style lag detection, and it contains standard DTVEM as a special case.

Understanding student errors in the programming is a cornerstone of programming education, yet obtaining a representative set of student errors for any newly designed task remains slow and costly, since authentic submissions only accumulate after extensive classroom deployment. This paper explores whether large language models (LLMs) can serve as scalable proxies for students by simulating realistic logical errors in code submissions. Using the CodeWorkout dataset of 74,000+ unique student Java submissions across 37 problems, we evaluate five LLMs under three mainstream prompting strategies: Input-Output (IO), Chain-of-Thought (CoT), and iterative Self-Refine. We assess performance along two key dimensions: diversity (the range of distinct error patterns) and alignment (alignment with authentic student mistakes), and examine how these vary by struggling level of programming tasks. Our quantitative findings reveal that while all models generate diverse errors, their alignment to human submissions diverges: Claude Sonnet 4 achieves the most balanced performance. In addition, we conducted a blinded expert annotation study (N = 401) comparing synthetic and authentic errors. This qualitative analysis confirms that the generated errors are functionally indistinguishable from authentic student errors. Moreover, higher-struggling-level problems elicit more diverse but less student-like errors. These results highlight trade-offs in using LLMs to simulate human learners and suggest design considerations for integrating synthetic errors into teachable agents, intelligent tutoring systems, and large-scale learning analytics.

We introduce Adelic operation-preserved embeddings (AOE), a training-free representation that captures both a number's real value and its modular (p-adic) signatures. This construction preserves additive and multiplicative structure by design, turning numerical input into embeddings that "speak in the language of mathematics." Unlike prior approaches that rely on task-specific retraining, AOE is plug-and-play and drops seamlessly into existing architectures. On algebraic combinatorics benchmarks, it delivers consistent gains including the first-ever perfect accuracy on the Weaving Pattern task-while suggesting a principled path forward for overcoming the long-standing "number problem" in AI.

Graphical User Interface (GUI) agents are increasingly used to automate complex computer tasks across applications, websites, and operating systems. To improve their reliability, recent work has introduced experiential memory, where agents retrieve prior trajectories to guide decision-making in similar states. More recent approaches further extend this idea to visual memory by storing and retrieving screenshots from past interactions, providing agents with richer contextual information than text-only memories. However, the effect of visual memory in GUI agents remains insufficiently understood: it is unclear which failures visual memory mitigates, or which failures it exacerbates. To systematically analyze the effect of visual memory, we introduce a taxonomy of four GUI agent failures (i.e., cognitive failure, visual state misunderstanding, hidden operation blindness, and grounding error) that map to distinct stages of the perception-reasoning-action pipeline. We find that prepending full-image memory has a divergent effect on the failure distribution: it reduces state-level failures but worsens action-level ones, and increases hidden operation blindness and grounding error. Motivated by this finding, we propose Action-Grounded Visual Memory (AGMem), an action-grounded memory framework for GUI agents. The core idea of AGMem is to store image crops that capture the local GUI region closely related to a successful action or a recovery, rather than storing full screenshots. Experiments on OSWorld show that AGMem improves task success rates by 33.3 % over full-image memory. These results demonstrate that AGMem is an effective representation for visual memory in GUI agents.

We study the sample complexity of learning in average-reward weakly-coupled Markov decision processes (WCMDPs) and Restless Bandits (RBs) under a generative model. Naive reduction to a tabular MDP leads to high complexity bounds as the state-action space is exponentially large in the number of arms $N$. By exploiting the weakly coupled structure, we show that near-optimal policies can be learned with sample and computational complexities that are polynomial in $N$. Specifically, we analyze the plug-in approach, which applies an efficient planning algorithm to an empirical model estimated from data. For fully heterogeneous WCMDPs, we establish the first finite-sample PAC guarantee with polynomial complexity and an $O(1/\sqrt{N})$ optimality gap. For homogeneous RBs, we further prove that a smaller optimality gap is achievable under mild structural assumptions. A primary technical contribution of our work is a novel Lyapunov-based analysis framework. Unlike classical approaches that rely on the difficult-to-control bias function, our framework uses an explicitly constructed Lyapunov function along with a drift transfer technique between the true and empirical models. A key step of independent interest in our framework is a fine-grained perturbation analysis for the underlying linear programming (LP) relaxation, which provides a general tool for analyzing LP-based policies and weakly-coupled systems.

Conventional RGB cameras have been widely used in multi-object tracking due to their ability to capture rich appearance and semantic information. However, their performance is often degraded under complex real-world challenges, such as motion blur, low illumination, and overexposure. Bio-inspired event cameras offer high temporal resolution and high dynamic range, providing complementary cues under extreme scenarios. Nevertheless, RGB-event multi-object tracking remains underexplored due to the lack of large-scale and well-annotated datasets. To address this issue, we propose FEMOT, a large-scale RGB-event multi-object tracking dataset that covers diverse real-world scenarios and 14 challenging attributes. With both RGB and event data as well as high-quality annotations, FEMOT provides a reliable platform for systematically evaluating RGB-event multi-object tracking methods. Based on FEMOT, we retrain and evaluate over ten strong trackers, thereby establishing a comprehensive benchmark for future research. Furthermore, we propose FEMOTR, a multimodal tracking framework that decouples RGB and event features and fuses them in the frequency domain, thereby effectively exploiting their complementary characteristics for robust object localization and identity association. Extensive experiments on FEMOT and DSEC-MOT datasets demonstrate the effectiveness of the proposed method. The source code and benchmark dataset have been released on https://github.com/Event-AHU/FEMOT.

Determining whether one concurrent operation completed before another began is a fundamental prerequisite for reasoning about the correctness of concurrent systems. We formalize this challenge as the Causal Observability Problem (COP): assign timestamps to the observable boundary events of a concurrent execution, invocations and responses, that faithfully reflect real-time operation order. A solution is complete if it never misses a genuine precedence, and sound if it never reports a spurious one. We prove that a strongly consistent solution, one that is simultaneously complete and sound, is unachievable at the observable boundary. We then show that the placement of instrumentation events relative to operation boundaries deterministically governs what a monitor can guarantee: internal placement yields completeness, external placement yields soundness, and neither achieves both. This dichotomy holds independently of the underlying timestamp mechanism. We instantiate this framework with three non-blocking implementations of a Causal Monitor object: FAInc (centralized atomic counter), Striped (decentralized counter), and Collect (iterative register snapshot). FAInc and Striped are linearizable; Collect is only quiescently consistent. Despite this internal consistency gap, we prove that all three provide identical COP guarantees: placement alone determines observable behavior. We validate these claims empirically on a 64-core NUMA architecture, showing that Striped matches Collect in throughput while preserving linearizability, resolving the cache-contention bottleneck of FAInc at high thread counts.

Rapid post-event landslide mapping is essential for disaster response but remains difficult to automate due to extreme class imbalance. This study evaluates whether Clay v1.5, a Geo-Foundational Model (GFM), can improve pixel-level landslide segmentation on the Landslide4Sense (L4S) benchmark, which contains 3,799 training chips with 14 Sentinel-2 and terrain bands and approximately 2% positive pixels. We compare three strategies: Clay as the primary encoder with multi-scale residual terrain fusion, a U-Net backbone augmented with Clay semantic context at the bottleneck, and a standard U-Net baseline. The hybrid U-Net + Clay model with two-stage Low-Rank Adaptation (LoRA) achieved the best test F1 of 64.5 +/- 1.8% over three seeds, surpassing the Clay-only backbone (55.2 +/- 3.6%) and the U-Net baseline (59.9%). Clay as a standalone encoder underperformed the U-Net due to the absence of multi-scale skip connections, but its pretrained representations consistently improved performance when injected as auxiliary context. These findings suggest that GFMs are most effective for landslide detection when they complement spatially detailed convolutional architectures rather than replace them.

We propose a spectral learning method for stochastic nonlinear dynamical systems represented with embedded latent transfer operators in deep feature spaces. We instantiate the method as Deep Spectral Encoder (DSE), an operator-based latent state-space model in which a time-invariant neural encoder implements learnable nonlinear feature maps from observations, and these features define Markovian latent states whose temporal evolution and observation mapping are described by the transfer and observation operators, respectively. Functional canonical correlation analysis in a learnable Galerkin-projected feature space provides state coordinates from past and future observations, and the two linear operators are estimated on the state coordinates as ridge-regularized closed-form solutions that coincide with Galerkin projections of the associated covariance operators. On this representation, we generalize sequential Bayesian filtering and Koopman spectral mode decomposition in feature space. Experiments on several scenarios show stable and superior performance with sequential Bayesian filtering and dynamic mode decomposition baselines even under noise and partial observability.

Existing studies reveal that current backdoor defenses exhibit limited robustness and often fail against specific types of attacks. More concerningly, prevailing safety tuning strategies tend to provide only superficial safety protection, as they fall short of completely eliminating the backdoor effects. In this work, we present a novel formulation of backdoor learning and unlearning as a sequential, three-stage process from a continual learning perspective. Within this framework, we formally define complete backdoor unlearning and further derive the necessary conditions for achieving it based on the mechanism of catastrophic forgetting. Guided by these insights, we propose Blind Inversion-Backdoor Adversarial Unlearning (BI-BAU), which formulates the generation of adversarial examples satisfying the unlearning conditions as a blind inversion problem. We solve this by integrating the bi-level optimization process of adversarial training into an Expectation-Maximization (EM) algorithm framework to optimize the maximum a posteriori (MAP) objective. Furthermore, BI-BAU is extended to untargeted adversarial scenarios with unknown target classes, as well as to multi-modal contrastive learning tasks, enhancing its applicability to real-world deployment scenarios where pre-trained models may be compromised. Extensive experiments demonstrate that our method exhibits general applicability across a wide spectrum of backdoor attacks and can effectively and thoroughly eliminate the backdoor effects from a backdoor model.

Real-temperature topological magnetic dynamics in functional materials is governed by coupled lattice and spin evolution, yet remains inaccessible to predictive simulation at device-relevant scales. As a flagship example, thermally driven helix-to-skyrmion transformation in FeGe requires atomistic resolution, explicit lattice motion, and micrometer-scale domains to resolve device-scale topological texture formation. We combine a spin-constrained density-functional-theory-trained neuro-evolution potential with a structure-preserving spin-lattice integrator within one machine-learned framework. Architecture-specific optimizations, kernel fusion, SVE2 vectorization, and NUMA-aware data layout deliver a seven orders-of-magnitude speedup over prior spin-aware methods. Deployed on LineShine exascale supercomputer, the full application scales to 12.45 million CPU cores with 89.7% weak-scaling efficiency, enabling simulations of 1.34 trillion atoms and an equal number of spins while reaching 48.5 PFLOPS in double precision. The simulations directly resolve real-temperature skyrmion nucleation and reorganization at previously inaccessible scales, establishing a new regime for predictive simulation of coupled spin-lattice topological magnetic dynamics.

Accurate pediatric brain tumor segmentation remains challenging due to limited annotated data, heterogeneous imaging phenotypes, diffuse tumor boundaries, and class imbalance across tumor subregions. Here, we present a two-stage deep learning framework for improving multi-modal pediatric brain MRI segmentation and clinical interpretation. First, we evaluate 3D Res U-Net and Swin-UNETR baselines on BraTS-PEDs MRI scans, using four co-registered modalities to predict tumor core, whole tumor, and enhancing tumor regions. Second, we introduce diffusion-based refinement models conditioned on coarse Swin-UNETR predictions, including a 3D DDPM refiner and MedSegDiff. Conditioning substantially improves diffusion stability and performance, particularly for enhancing tumor boundary segmentation. Conditioned MedSegDiff achieves the strongest boundary agreement with the lowest HD95. Finally, predicted tumor volumes and representative segmentation overlays are integrated with a multimodal language model to generate structured radiology-style reports. Together, our results suggest that coarse-to-refined diffusion segmentation can improve pediatric tumor boundary delineation and support end-to-end interpretable AI-assisted neuro-oncology workflows.

Temporal online analytical processing (OLAP) analyzes past states of data whose values change over time. Such histories are naturally stored as interval histories, in which each row records the period during which a value remained valid. Because temporal analyses typically arrive in infrequent, intermittent bursts, serverless execution that launches functions only at query time offers a cost advantage over always-on clusters. Splitting a computation that a single process performs as a whole across independent serverless functions, however, breaks correctness in two ways. A function may not receive the rows that determine the state of its time range, and naively summing partial results yields incorrect answers for duration-weighted and cumulative-threshold queries. Existing SQL engines and serverless analytics do not address both problems together. This paper presents Vivace, a serverless system for exact temporal OLAP over interval histories. Vivace resolves the two problems in separate stages. Before any query arrives, a pre-query layout step partitions the interval history, replicating boundary-crossing intervals so each function computes its range completely from a single file. At query time, a merge step combines partial results under operator-specific rules. Associative aggregates merge intermediate values, and ranking re-orders candidates within each time range. We prove that this partitioned execution matches single-process computation up to canonical form. Evaluated on AWS Lambda with real-world datasets, Vivace reduces latency and monetary cost by up to 82% and 84%, respectively, against an equivalent SQL baseline that queries the history directly, demonstrating robust generality and efficiency.

Existing studies on gender bias in LLMs have largely focused on stereotypes, occupational associations, or explicit harmful outputs. In this work, we ask whether LLMs apply consistent response standards to the same negative behavior under matched male-actor and female-actor conditions. We introduce GAMA-Bench, a gender-mirrored benchmark of 1,298 scenarios covering intimate relationship and public social conflicts. It constructs gender-neutral misconduct templates through controlled grids and cross-model review, then compiles them into paired first-person prompts with matched actor-gender and role-reference variations. We further design a structured response-framing protocol to measure how models allocate punishment, empathy, escalation, instruction, and blame. Experiments on 10 representative LLMs reveal a consistent male-disadvantaging asymmetry: male actors receive more punitive, escalatory, and blame-centered framing, whereas female actors receive more therapeutic and empathy-oriented framing for the same misconduct. Further analyses show that this pattern persists across model families, scenario tracks, model scale, and explicit thinking-style reasoning. The official code is available at https://github.com/xufeiqiong/GAMA-Bench.

Large Language Model (LLM) coding agents have achieved strong results on software engineering tasks, yet repository exploration remains a major bottleneck: locating relevant code consumes substantial token budget and pollutes the agent's context with irrelevant snippets. In most agents, the same model explores the repository and solves the task, leaving exploratory reads and searches in the solver's history. We present FastContext, a dedicated exploration subagent that separates repository exploration from solving. Invoked on demand, FastContext issues parallel tool calls and returns concise file paths and line ranges as focused context. FastContext is powered by specialized exploration models spanning 4B--30B parameters. We bootstrap them from strong reference-model trajectories and refine them with task-grounded rewards for broad first-turn search, multi-turn evidence gathering, and precise citation generation. Across SWE-bench Multilingual, SWE-bench Pro, and SWE-QA, integrating FastContext into Mini-SWE-Agent improves end-to-end resolution rates up to 5.5\% while reducing coding-agent token consumption up to 60\%, with marginal overhead. These results show that repository exploration can be separated from solving and handled effectively by specialized models. Code and data: https://github.com/microsoft/fastcontext

Coding agents powered by large language models have demonstrated strong performance on software engineering tasks. Yet most agents consume repositories almost entirely as text, which differs from how human developers use visual structure such as folder hierarchies and dependency relationships to orient themselves in large codebases. With multimodal large language models (MLLMs), it is an open question whether agents can effectively benefit from visual representations of repositories. This paper presents the first systematic empirical study of visual repository representations for LLM-based agents on repository-level issue resolution. We evaluate four recent multimodal models. Our results show that a strictly vision-only setup degrades accuracy and increases token cost, because agents lack sufficient symbolic detail and compensate with repeated visual queries. In contrast, integrating visual graphs of repository structure as a supplementary modality alongside standard text interfaces helps agents understand structure more efficiently: input token consumption decreases by up to 26% while issue-resolution accuracy is maintained or improved. Visualization is most useful during fault localization and when the agent autonomously controls exploration depth. These findings point to a practical hybrid text-and-vision design for next-generation coding agents.

Adversarial conditions such as paraphrasing and targeted style transfer sharply degrade the accuracy of machine text detectors. A document, however, carries multiple complementary signals (e.g., stylistic features, likelihood and rank-order features, and structural features), and an attack that suppresses one may leave others intact. While a parametric classifier can learn to combine these features given sufficient supervision, classifiers are prone to making confidently incorrect predictions when the distribution shifts (e.g., novel attacks or unseen language models). To address this, we propose a multi-view, non-parametric detection framework that extracts complementary feature views from the same document and aggregates per-view evidence through a Gaussian process ensemble. By aggregating evidence across views, an adversary must simultaneously defeat multiple independent axes of detection, substantially raising the cost of evasion. The Gaussian process formulation additionally provides calibrated probabilities and principled abstention on out-of-distribution inputs, supporting reliable deployment in high-stakes settings. We evaluate on three benchmarks spanning diverse generators and attacks: the DetectRL and RAID benchmarks, and the PAN2025 shared task and demonstrate that our multi-view detector maintains strong performance under the considered attacks, outperforming existing approaches against held out attacks.

Quantifying the influence of hybrid aleatory and epistemic uncertainties on high-dimensional system responses remains a major challenge in global sensitivity analysis (GSA). Existing Hilbert--Schmidt Independence Criterion (HSIC)-based approaches are primarily restricted to single-output settings and lack a rigorous decomposition of heterogeneous uncertainty sources and their interactions. To address this limitation, a novel double-space tensor-product RKHS framework is proposed for sensitivity analysis under hybrid uncertainty. By constructing factorized kernels over both the latent input space and the multidimensional output space, a concurrent double Möbius inversion is derived to orthogonally decompose the global dependence measure into pure aleatory effects, pure epistemic effects, and their interaction contributions. The resulting dimension-wise sensitivity indices preserve the uncertainty attribution structure across all output dimensions. To satisfy the independence assumptions required by the decomposition, an auxiliary-variable representation based on the inverse probability integral transform is introduced, enabling the treatment of hierarchical uncertainties and Copula-induced correlations within a unified latent space. A fully vectorized single-loop implementation is further developed to avoid the computational burden of nested Monte Carlo simulation. Statistical significance and estimation uncertainty are quantified through permutation testing and Bootstrap confidence intervals. Numerical studies on a modified multi-output Ishigami function and an aerodynamic pressure-field problem demonstrate the accuracy, scalability, and practical applicability of the proposed framework.

Long-context language modeling requires not only extending context windows but maintaining coherent understanding of entity states and relationships across thousands of tokens -- a challenge that semantic similarity alone cannot address. KGERMAR addresses this by constructing dynamic, context-specific knowledge graphs from input text during inference, enabling domain-adaptive retrieval that leverages both semantic similarity and explicit entity relationships. The framework performs real-time entity and relation extraction to build contextual knowledge graphs, then integrates graph-structural embeddings with textual semantics through a multi-component memory architecture. Three memory banks -- contextual, semantic, and structural -- are maintained with retrieval signals fused via learned weights to capture both surface-level semantics and deeper relational patterns. Evaluated on SlimPajama (84.7K training examples), WikiText-103 (4,358 examples), PG-19 (100 examples), and Proof-pile (46.3K examples), KGERMAR achieves up to 8.5\% lower perplexity and 2--2.5x better memory efficiency than memory-augmented baselines across context lengths from 1K to 32K tokens, with superior in-context learning performance across five NLU tasks. The dynamic knowledge graph construction approach advances memory-augmented language modeling by enabling domain-specific knowledge representation that adapts to input contexts rather than relying on fixed knowledge bases.

Sparse autoencoders (SAEs) are typically trained to reconstruct the \textbf{entire} residual stream through a sparse dictionary, implicitly assuming that all activation content is amenable to sparse, monosemantic decomposition. We question this assumption and hypothesize that activations contain a low-rank, dense component that is computationally important to the model yet inherently unsuitable for sparse representation, which serves as a major source of the persistent dense latents widely observed in trained SAEs. To test this, we add a small rank-$r$ linear bottleneck in parallel with standard SAEs (BatchTopK and Matryoshka), allowing dense structure to be absorbed before sparse reconstruction. On Gemma-2-2B layer 12, a rank-24 bottleneck reduces dense latent count by up to 84\% while improving sparse probing and targeted probe perturbation on both architectures at matched sparsity. The absorbed component is (i) \textbf{structurally identifiable} as the top principal components and outlier dimensions; (ii) \textbf{causally necessary}, with removing it raising next-token cross-entropy by 7.5$\times$, far exceeding the 2.8$\times$ from removing the geometrically near-identical top-24 PCA directions; and (iii) \textbf{redundantly encoded by sparse dictionaries}, with ablating 787 maximally aligned sparse features raising cross-entropy by only 2.9$\times$ and ablating 2,048 topic-aligned features leaving MMLU topic classification virtually unchanged, whereas removing the scaffold drops it from 98.7\% to chance. Together, our findings identify a compact, semantically informative and causally important component of residual stream activations (which we term a \textbf{computational scaffold}) that standard sparse dictionaries represent inefficiently, suggesting that the scope of sparsity-based interpretability methods warrants careful re-examination.

As language models take integrated roles across many domains, the response of LLMs to user pushback becomes a critical alignment property. Yet many existing evaluations treat compliance as unidirectional, measuring whether models resist pressure but not whether they resist it selectively. We introduce Compliance Asymmetry (A = BCR/HCR), a bidirectional diagnostic that compares beneficial output change under helpful nudges with harmful change under misleading nudges. Across 9 models and 972,000 nudge-condition responses, we find that this selectivity differs in factual and moral judgments: models follow helpful nudges more than harmful ones on factual questions (A = 1.58), but follow both directions at nearly identical rates on moral questions (A = 1.04). This phenomenon persists across model families, capability levels, and nudging types. Interestingly, we also find that chain-of-thought prompting amplifies helpful and harmful compliance together, while identity-based prompting suppresses both by nearly identical margins. These results identify direction-blind moral compliance as a distinct failure mode in current LLMs and suggest that alignment should target directionally calibrated updating rather than lower compliance alone.

Identifying conditions that a certain drug takes therapeutic effect on a target disease is crucial for clinical decision-making support. However, most existing biomedical information extraction methods have focused on identifying only relations between drugs and diseases, while largely overlooking the context-specific conditions where such relations can apply. To address this problem, we introduce the task of applicability condition extraction for therapeutic drug--disease relations from biomedical research literature. We create the first dataset that has manually annotated triples of drugs, diseases, and applicability conditions on biomedical paper abstracts with 1,119 drug-disease pairs. Using this dataset, we systematically evaluate the performance of a range of existing methods. In addition, we propose a new method that enhances LoRA to consider relations between drugs and diseases. Our method consistently outperforms strong baselines across different evaluation settings. The source code and dataset of this paper can be obtained from: https://github.com/guantingluo98/Drug-ACE

Streaming speaker diarization is crucial for time-critical medical dispatch, but deploying it on resource-constrained hardware requires smaller, faster models. Using SIMSAMU, a dataset of simulated medical-dispatch conversations, we evaluate streaming behavior before compressing the segmentation model with pruning and low-bit quantization. We characterize performance across a range of streaming latency budgets and find that additional buffering is not consistently beneficial, while very low-latency operating points can substantially degrade performance. Our study shows that model compression trades performance for memory footprint, and we highlight an operating point where FP16 reduces model size by half with essentially unchanged real-time factor, at a cost of a 40\% relative DER increase against the baseline. This work characterizes the trade-offs for real-time deployment and contributes to speech technology that can enable reliable human communication in time-critical contexts.

Constrained MDPs (CMDPs) are a widely adopted framework for incorporating safety into RL agents; however, the framework does not support risk-sensitive constraints. This can be problematic: For example, CMDPs allow for optimal solutions that, in order to satisfy the risk-neutral constraints, mix infrequent catastrophic behaviors and frequent, overly conservative ones. Moreover, prior empirical results suggest that enforcing stricter, risk-sensitive constraints can improve performance even under risk-neutral evaluation. The natural framework to incorporate risk-sensitive constraints is utility-constrained MDPs (UCMDPs), but no practical solutions for this problem existed. In this work, we introduce a simple yet powerful methodology for UCMDPs and constrained RL. Besides allowing for risk-sensitive constraints, our framework does not require us to fix constraint limits in advance of training the agent, provided that a sensible range is known. This increases policy flexibility and, in practice, allows for adjustments to these limits at no extra training cost. Besides benefiting from the generality of the framework, our agent shows strong performance in practice, consistently matching or outperforming existing baselines in several Safety Gymnasium benchmark tasks.

In small-batch scientific deployments, labeled target outcomes may be too scarce for reliable shift estimation even when unlabeled target inputs are available. We address the complementary setting where the practitioner has a pre-specified label-shift correction from domain knowledge and asks whether incoming labeled outcomes support it. We show that the per-observation likelihood ratio between a label-shift-corrected predictive and the source predictive is a conditional e-value, so its running product is a nonnegative martingale and Ville's inequality yields an anytime-valid confirmation rule. The log martingale equals the cumulative negative log-predictive density (NLPD) gap between the source and the corrected predictive, converting routine model monitoring into a formal sequential test. Rejection means the incoming data support the posited correction relative to the source predictive, but it is not a precise estimate of the degree of shift. Closed forms are available for GP sources with Gaussian label-shift ratios. GP regression simulations validate Type I control, finite-sample power, miscalibration sensitivity, and the small-batch advantage of a reliable prior over label-based re-estimation.

Agentic browsers integrate autonomous AI agents into web browsers, enabling users to accomplish web tasks through natural-language instructions. The same-origin policy (SOP) is a fundamental browser security mechanism that prevents unauthorized automated cross-origin data flows induced by scripts. However, whether SOP remains effective in agentic browsers is an open question that has not been systematically studied. In this work, we bridge this gap. We first observe that an agentic browser can itself serve as an automated channel for cross-origin data flows, potentially leading to SOP violations. To investigate this phenomenon, we construct SOPBench, a benchmark for evaluating SOP violations in agentic browsers. Our evaluation shows that existing agentic browsers frequently violate SOP, both in benign settings and under attacks. To address this problem, we propose SOPGuard, an SOP enforcement mechanism tailored to agentic browsers. We implement SOPGuard in BrowserOS, an open-source agentic browser. Extensive evaluations demonstrate that SOPGuard effectively enforces SOP while preserving utility and incurring only a small runtime overhead. Our code and data are available at https://github.com/wxl-lxw/BrowserOS-SOPGuard.

Optimal Transport has become recently a powerful method for domain adaptation by aligning source and target distributions. We study a supervised domain adaptation problem where source and target domains are related by a rotation or a translation or a homothety in $\mathbb{R}^2$. We prove that the optimal transport map recovers the underlying map when using a $p-$norm cost with $p \geq 2$. Based on this insight, we develop a method combining $K-$means and optimal transport to estimate the underlying map, enabling adaptation of linear regression models when target data is scarce. Simulations demonstrate improved performance over baseline methods. Rather than relying on highly expressive deep learning architectures, we focus on classical machine learning models to emphasize interpretability and theoretical insight. This perspective allows us to explicitly characterize the role of optimal transport in recovering geometric transformations such as rotations, translations, and homotheties. Our contributions include a theoretical result linking optimal transport and rotations, translations and homothecies in $\mathbb{R}^2$, and a practical method for adaptation in linear regression offering both conceptual clarity and applied value in domain adaptation tasks in this space.

LayerNorm-based GNNs routinely erase the topology signals (degree, centrality, $k$-core) that node-selection policies should depend on, but the literature has not located where in the residual block the erasure happens. We answer that question: a positive per-node scalar inserted before LayerNorm is divided out up to a stabilizer term, while the same scalar inserted after LayerNorm reaches the score head as representation magnitude. The surviving slot is the post-LayerNorm position. We instantiate it with PostDeg, a parameter-free post-LayerNorm inverse-degree scale, and pre-register four falsifiers (graphwise scalars, extra LayerNorm, expressive same-slot capacity, backbone-agnostic source) that would reject the rule. PostDeg gains $+3.5\%/+2.5\%/+5.6\%$ over the LN backbone on influence maximization, network dismantling, and maximum independent set, with $10/10$ paired-seed wins per task; none of the four falsifiers fires. The takeaway is that placement, not parameterization, carries the gain -- a small invariance check that generalizes to any positive topology scalar in any normalized residual stack.

Vision-Language-Action (VLA) models have shown strong potential for end-to-end autonomous driving by jointly modeling visual perception, language reasoning, explainability and action prediction. However, their large vision-language backbones and reasoning modules introduce substantial inference latency and thereby prevent their deployment in the unforgiving reality of the road networks. We propose RT-VLA, a lightweight, distilled VLA model that transfers the driving and reasoning capabilities of the state-of-the-art SimLingo model into a compact student through multi-level supervised distillation. RT-VLA preserves language-based reasoning and supports post-hoc explanation through offline language analysis of safety-critical driving moments without adding latency to real-time control. Compared to the SimLingo teacher, RT-VLA maintains competitive closed-loop driving and language reasoning performance while reducing inference time by 44.8X in vision-only mode and 7.9X in vision+language mode. These results suggest that supervised distillation is a practical approach for building real-time, explainable VLA-style autonomous driving models.

Blockchain consensus mechanisms based on Proof-of-Work consume significant energy, with Bitcoin alone estimated at approximately 150 TWh per year. Proof-of-Space reduces this cost by replacing repeated computation with storage, but plot generation remains bottlenecked by CPU hashing throughput. Prior work on VaultX demonstrated a high-performance CPU-based Proof-of-Space plotter using multi-threaded Blake3 hashing, achieving plotting speeds 4 to 50x faster than Chia depending on hardware configuration. In this paper, we present VaultxGPU, a GPU-accelerated extension of the VaultX plotter that offloads the Blake3 hashing pipeline to the GPU using custom kernels. We implement the plotter in both CUDA for NVIDIA hardware and SYCL for AMD and Intel GPUs, keeping Table 1 entirely in GPU VRAM and fusing the sort and match stages into a single kernel to minimize data movement. We evaluate VaultxGPU across K-values 27 through 31 against CPU baselines. Our SYCL GPU implementation achieves a 59.2x speedup over a single-threaded CPU baseline, completing a K=31 plot in 45.4 seconds compared to 2688 seconds, and outperforms even the best 384-thread CPU configuration. These results confirm that GPU acceleration is the correct direction for scaling Proof-of-Space plotting beyond what CPU parallelism can achieve.

Maritime situational awareness often relies on Automatic Identification System (AIS) transmissions to track vessel movements. However, in operational or conflict scenarios, these data may be unavailable due to signal loss, deliberate deactivation, or intentional spoofing. In such conditions, synthetic aperture radar (SAR) imagery becomes a critical sensing alternative for wide-area maritime monitoring, despite providing only static scene snapshots. This work introduces HARBOR (Heading Analysis and Reconstruction from Behavioral Observation and Radar), a complete pipeline for transforming a single SAR image into predictive motion information without requiring any auxiliary data source at inference time. The method begins with SAR image preprocessing to enhance and segment vessel candidates, followed by automatic detection, size-based classification, and heading estimation using skeleton geometry and local intensity patterns. AIS data are used exclusively during an offline calibration phase to derive vessel-type-dependent motion parameters, which are then applied to generate probabilistic heatmaps of candidate future vessel positions. A case study using real COSMO-SkyMed SAR imagery demonstrates the pipeline on a maritime scene in southern Brazil, showing its ability to extract motion tendencies and generate probabilistic projections of vessel positions in data-denied environments.

Determining the crystal structure of a material from its powder X-ray diffraction (PXRD) pattern is a central challenge in materials science. PXRD is an accessible and widely used characterization technique, yet recovering the atomic structure from diffraction data requires solving an underdetermined inverse problem due to the loss of phase information. Generative modeling can provide a prior over atomic structure and learn the mapping from PXRD patterns to crystal structures via simulated structure-spectrum pairs. We present XRDiff, a diffusion model that recovers crystal structures from PXRD given either the stoichiometry or, in a more challenging setting, the elemental constituents and total number of atoms in the unit cell. We evaluate on datasets where each stoichiometry has multiple polymorphs and all polymorphs of a given composition are held out together, ensuring that high performance reflects genuine use of the diffraction signal. XRDiff achieves strong structure recovery rates on simulated benchmarks, indicating that the model learns a spectrum-to-structure mapping precise enough to differentiate between polymorphs. To address generalization to experimental data, we compare a full-spectrum encoding against an encoding based on peak descriptors. The peak-based encoding generalizes substantially better, outperforming even a model trained on full spectra with augmentations fitted to the experimental noise distribution. These results demonstrate that representations robust to the noise and artifacts present in real-world PXRD offer a practical and scalable path toward closing the simulation-to-experiment gap, enabling zero-shot crystal structure solution from experimental PXRD with full or partial chemical composition input.

Recent work has demonstrated that coding agents can formalize entire advanced mathematics textbooks in Lean 4, yet existing efforts concentrate on branches of mathematics already well-represented in mathlib and measure success solely through kernel acceptance. We address both limitations by applying a coding agent to formalize Numerical Methods for Ordinary Differential Equations, a textbook in numerical analysis that is largely absent from mathlib, stressing the agent's capacity to develop new theory from scratch. We further introduce a systematic, reproducible three-dimensional framework for evaluating the quality of agent-produced formalizations beyond compilation: semantic correctness, Mathlib reuse, and cross-file reuse via LLM-as-judge methods. Applying this framework to our own formalization and to the released outputs of RepoProver and M2F, we uncover recurring unfaithful formalization patterns, including incomplete multi-part statements, added weakening hypotheses, and parameter restrictions, that kernel acceptance entirely obscures. Our results suggest that compilation-based metrics substantially overstate formalization quality, and we provide a reproducible audit methodology to support more rigorous evaluation of future autoformalization systems.

AI coding agents have rapidly transformed software engineering, powering widely used interactive coding assistants. Despite their interactive real-world use, existing benchmarks evaluate them as fully-autonomous systems. In this work, we introduce Dialogue SWE-Bench, an automatic benchmark dataset for evaluating the ability of coding agents to resolve real-world software engineering problems through dialogue with a user. We design a novel, persona-grounded user simulator to support our task evaluation, and augment our task evaluation with automatic evaluations of dialogue quality. We also propose a new schema-guided agent, aimed at improving the dialogue capabilities of off-the-shelf coding agents, which improves over strong baselines by 3-14%. Our results indicate that better coding models do not always correspond to better dialogue models, suggesting that dialogue capability is a distinct and currently understudied dimension of coding agent performance.

LLM-based Agents are becoming increasingly capable and widely deployed, creating growing incentives for adversarial misuse in the real-world. A key emerging threat is Decomposition Attacks \cite{glukhov2024breach, jones2024adversaries} in which a harmful task is broken into simpler, benign subtasks that evade safety mechanisms when executed separately but cumulatively fulfill the malicious intent. Although recent benchmarks assess agent safety in multi-turn and multi-tool-use settings, they do not explicitly capture this form of decompositional misuse and may not represent realistic adversarial execution flows. To this end, we introduce DeCompBench, a benchmark designed specifically to evaluate agentic safety under decomposition attacks. DeCompBench is created with a decomposition-by-design principle using a graphical framework and enables harmful task decomposition into individually benign and executable subtasks with realistic workflows. Our experiments using a custom decomposer show that state-of-the-art agents exhibit high refusal rates on monolithic harmful tasks, but significantly lower refusal rates on their decomposed variants, while often inadvertently fulfilling the adversarial objectives. These findings underscore the need for safety evaluations against decomposition attacks and corresponding defenses. Our dataset is publicly available and can be found at https://huggingface.co/datasets/decompositionbench/DeCompBench.

A crucial aspect of linguistic capability is the ability to trade off between stored representations and abstract knowledge: one must retrieve learned representations, but also generate novel ones by applying productive rules. While recent work has examined abstract knowledge in language models, holistic storage of multi-word units has received far less attention. We probe internal representations in text-based LLMs and an ASR model, testing whether V+up phrasal verbs develop distinct representations as a function of frequency and predictability. All models show evidence of holistic storage driven by frequency and predictability, further supporting usage-based theories of language.

The likelihood ratio framework is widely recognized as the logically and legally sound basis for evidential analysis across forensic sciences, and its importance is increasingly acknowledged in analyses of authorship in textual evidence. To date, however, its application has been confined to English-language texts. Meanwhile, authorship attribution has traditionally relied on a diverse array of stylometric features, even as the rise of pre-trained large language models enables new contextual-embedding approaches. Combining these diverse approaches through fusion promises enhanced performance, yet it has not been applied to integrate stylometric-feature systems with embedding-based systems within the likelihood ratio paradigm. This study is the first to apply likelihood ratio-based forensic text comparison to Japanese digital texts, using ~1,000-character excerpts from blogs, to 1) evaluate system performance and likelihood ratio magnitudes and 2) assess the impact of fusing stylometric-feature systems with embedding-based systems. The results demonstrate that the fused system maintains excellent calibration while 1) increasing consistent-with-fact likelihood ratio magnitudes; 2) decreasing contrary-to-fact likelihood ratio magnitudes and 3) improving overall discriminability. The best-performing fusion achieved a log-likelihood-ratio cost of 0.32484, illustrating both the feasibility of likelihood ratio framework for Japanese and the benefits of fusion across heterogeneous systems.

Recent alignment-free non-autoregressive (NAR) text-to-speech (TTS) models formulate synthesis as a conditional infilling task, bypassing explicit duration predictors and external aligners. When speech is represented with neural codec tokens, the infilling problem becomes discrete, making Discrete Flow Matching (DFM), a Continuous-Time Markov Chain (CTMC) framework for discrete generation, a natural fit. However, inference-time control for stable low-step conditional infilling remains underexplored. We propose Mask, Sample, Revise, an inference-time CTMC stack for alignment-free DFM-TTS. The stack combines predictor-free guidance to strengthen text conditioning, prompt-matched conditional coupling to align the probability path with the acoustic prompt, and SC-ReMask, a schedule-constrained remasking mechanism that introduces token-to-mask transitions so early de-masking decisions can be revised. These components require no post-hoc fine-tuning and operate in a single tau-leaping sampler. Controlled ablations show that this stack improves intelligibility and robustness in the low-NFE prompted setting, outperforming unguided and guidance-only samplers with substantially more steps.

Evolutionary optimization of spiking neural networks (SNNs) becomes increasingly difficult as task complexity grows because they must search a combined topology--parameter space that grows super-exponentially with network size. We address this scaling challenge through a co-evolutionary ensemble framework in which a population of candidate SNNs is evolved with fitness defined by each network's marginal contribution to group performance. Grounded in cooperative game theory and difference evaluation functions from multiagent systems, this credit assignment rewards networks that consistently improve ensemble performance and penalizes redundancy, encouraging complementary specialization during evolution rather than relying on post-hoc combination of independently trained networks. We evaluate the approach on classification, regression, and control tasks under $μ$Caspian neuromorphic hardware constraints. Co-evolved ensembles achieve statistically significant improvements over both single-network evolution and post-hoc ensembles across all tasks, with the most pronounced gains in control, where standard evolution fails to discover effective policies and co-evolution enables a qualitative transition to near-optimal performance.

Decision trees are one of the fundamental tools in statistical learning due to their interpretability, flexibility, and their ability to adapt to nonlinear structures. Among them, the Classification and Regression Trees, introduced by Breiman, Friedman, Olshen, and Stone in 1984, became one of the most influential algorithms and remains one of the most widely used methods for classification and regression problems. On the other hand, Bregman divergences, introduced by Lev Bregman in 1967 in the context of convex optimization, provide a broad family of loss functions that naturally generalize the squared Euclidean distance. This family includes, among others, the Kullback-Leibler divergence, the Poisson divergence, and the Itakura-Saito divergence, as well as several losses associated with distributions belonging to the exponential family. Moreover, Bregman divergences possess a rich geometric structure and deep connections with convex analysis and information geometry. In this work, we propose a generalization of the CART paradigm based on Bregman divergences, thereby obtaining a broader family of decision trees adapted to different statistical models and underlying geometries. Although algorithms such as CART or classical implementations such as rpart incorporate different impurity criteria, these are usually introduced in an ad hoc manner for each specific model. In contrast, the Bregman divergence approach provides a unified framework that allows these criteria to be derived and interpreted from common convex and geometric principles. Beyond the algorithmic construction, we also investigate theoretical properties of these trees. In particular, we study how properties of the generating convex function -- such as strong convexity or smoothness -- influence impurity gains between parent and child nodes, as well as stability and consistency properties of the estimator.

Sampling plays an important role in long-form language-model reasoning. Over thousands of decoding steps, small changes in the candidate token set can compound into different reasoning trajectories, stability profiles, and final answers. Existing truncation methods such as top-$p$, min-$p$, and fixed top-$nσ$ sampling improve over unrestricted sampling, but they rely on fixed thresholds that cannot adapt to changes in entropy, task difficulty, training stage, or generation budget. We introduce Adaptive Nucleus Truncation Sampling (ANTS), which extends top-\(nσ\) sampling from a fixed decoding rule into an adaptive rollout-control mechanism for long-form generation. ANTS selects standardized neighborhoods around the maximum logit before temperature scaling, adapts the truncation width using an entropy-conditioned controller, and retains a no-truncation fallback arm to stabilize training when truncation becomes unsafe. On a 33B-total / 4B-active sparse Mixture-of-Experts reasoning model, ANTS improves average performance over percentage-based benchmarks by +1.9, +3.8, and +5.2 points at 8K, 16K, and 32K generation budgets, respectively. The strongest gains appear on instruction following and mathematical reasoning, with IFBench improving by more than 10 points at 32K and AIME 2025 improving by 7 points. Code generation reveals an important budget interaction. On Codeforces, ANTS trails the baseline at 8K, but reverses this gap and substantially improves ELO at 16K and 32K. These results suggest that sampler design should be treated not just as a decoding hyperparameter, but as part of how we stabilize and scale long-budget reasoning.

This paper evaluates a signal-processing and supervised-learning pipeline for classifying SDSS DR17 astronomical spectra into stars, galaxies, and quasars. Each spectrum is represented by its measured flux and inverse-variance information, combining spectral shape with a wavelength-dependent reliability profile. After resampling onto a common logarithmic wavelength grid, the flux and inverse-variance vectors are standardized and separately compressed using principal component analysis. The resulting components are concatenated and used to train several classifiers. The best performance was obtained with the LightGBM gradient-boosting classifier, reaching $94.6\%$ accuracy and $92.1\%$ balanced accuracy on the test set.

"Integrative" solutions are widely praised but rarely defined: we lack an operational way to tell a genuine integration -- one that makes the world cheaper to describe -- from a tidy re-description. Building on the lineage that treats creativity and intelligence as compression, we give such a criterion for creative integration (CI): the resolution of a real conflict between A and B is CI if and only if, under a fixed description language, the description length strictly shrinks (C = L_pre/L_post > 1), with the reduction located in the conflict itself. We make the judgment decidable through four binary, conjunctive gates, and we fix its extension through a taxonomy of pseudo-integration that names and rejects the look-alikes. We back the criterion with a curated, multi-domain corpus and -- crucially -- validate it not by human inter-rater agreement but by four falsifiable tests it could fail: an independent computational check, discrimination against hard negatives, out-of-sample prediction, and description-language robustness; all pass with margin. The contribution is not "creativity is compression" but its decidability, discrimination, and corpus: on this account, what makes a move genuinely creative -- rather than merely novel -- is that it compresses a conflict, with novelty and value as downstream symptoms; whether all creativity is so constituted we state as an explicit conjecture. We claim only the sign of C-1; we judge, not generate. The result is a citable primitive for a broader program.

Learning with missing modalities is a fundamental challenge in multimodal robot learning, as real-world robotic systems often operate in environments with incomplete sensor data. Attention-based models are appealing for processing multimodal data because they can handle multiple modalities with a single backbone network. However, most multimodal models assume that all modalities are available during both training and inference, limiting their applicability in robotic perception and decision-making. In this paper, we introduce a multimodal model designed to handle missing modalities during both training and inference. The model is formulated as a conditional variational autoencoder (CVAE) and incorporates a transformer-based architecture that leverages attention mechanisms to learn a unified, fixed-dimensional representation, even when some modalities are missing. We show that our proposed model can be trained with missing modalities while approximating a robust representation of all modalities. We evaluate our approach on five multimodal datasets across two robot learning tasks: human trajectory prediction and robot manipulation forecasting. Experimental results demonstrate that our model effectively learns from incomplete data and is superior to prior multimodal fusion approaches.

Researchers and practitioners working with large language models face a fragmented landscape: local models are free and private but hardware limits the model size and context windows a researcher can use; institutional HPC centers offer powerful GPU resources at no marginal cost and keep data within institutional boundaries, but operate behind firewalls and are designed for batch jobs rather than interactive use; commercial cloud APIs provide frontier-model quality on demand but impose significant cost and data retention policies unsuitable for sensitive research data. No existing system unifies all three. STREAM (Smart Tiered Routing Engine for AI Models) addresses this gap with four contributions: (1) a three-tier routing architecture combining local, HPC, and cloud inference with a local LLM-based complexity judge; (2) a dual-channel HPC streaming architecture that separates the Globus Compute control plane (authentication and job dispatch) from a WebSocket relay data plane (token delivery), enabling sub-second TTFT (0.54 s median, 21.1x over batch mode's 11.40 s) through institutional firewalls without VPN or firewall rule changes, with end-to-end AES-256-GCM encryption ensuring the relay operator cannot read token payloads; (3) tier-aware context summarization that prevents long conversations from forcing simple queries onto expensive tiers; and (4) an HPC-as-API proxy mode that exposes HPC inference as an OpenAI-compatible endpoint callable from any standard client with no HPC expertise, a deployment pattern made practical only by the sub-second TTFT of contribution (2). Llama 3.2 3B achieves 85.1% free-tier retention on a 1,200-query benchmark spanning ten domains. Measured TTFT: 0.26 s local, 0.54 s HPC (relay), 1.68 s cloud.

The integration of artificial intelligence into human decision-making environments has introduced a previously undertheorized cost: the gradual surrender of human autonomy in exchange for access to information and computational assistance. Building on the Human Identity and Autonomy Gap (HIAG) framework, this paper advances a theoretical model of autonomy surrender as a measurable, cumulative process driven by cognitive bandwidth depletion. The model proposes three interacting mechanisms: the silent cost of AI assistance, in which autonomy is transferred incrementally and without awareness; the surrender threshold, beyond which reclaiming autonomous function becomes cognitively and psychologically difficult; and the recovery mechanism, which establishes the design obligation and the ethical responsibility accompanying deliberate human re-assumption of control. The paper argues that human re-entry into the decision loop is not a passive option but an active cognitive event requiring intentional bandwidth restoration. The design of AI systems must incorporate structured re-entry pathways, here termed recovery mechanisms, that preserve human agency while appropriately distributing responsibility. The model further predicts a terminal state, here termed preference inversion, in which functional dependence on AI assistance is experienced not as a deficit but as a preference, transforming the restoration of autonomy from a design problem into a cultural and political one. Implications are drawn for AI system design, governance frameworks, and human factors research.

Sierra Leone's agriculture operates with almost no data-driven decision support, and no published machine learning study has examined the country's crop yields. We ask whether rice yield can be forecast from data Sierra Leone currently has. Using 25 years of FAOSTAT production data (2000-2024) for nine major crops, we train XGBoost, Gradient Boosting, and Random Forest under a strict anti-leakage protocol with expanding-window walk-forward evaluation across seven held-out years, benchmarked against naive persistence. No model trained on crop statistics alone outperforms persistence. Augmenting with free satellite climate data (CHIRPS rainfall, NASA POWER temperature) reverses this result: a climate-only XGBoost reduces forecast error by one third (RMSE 284 vs 428 kg/ha), a gain that holds for a linear model and is robust to excluding the anomalous 2018 season. Early-season (May-June) rainfall is the dominant predictor, implying seasonal yield risk is observable months before harvest. No model anticipated the 2018 collapse, whose origins were institutional rather than climatic. We translate the findings into policy recommendations for Sierra Leone's Feed Salone Strategy, with a fully open-source pipeline.

Latent diffusion is a promising framework for scalable 3D molecular generation, but it requires a latent space that remains smooth, valid, and navigable beyond posterior samples. Existing molecular VAEs, however, are typically learned through reconstruction-based objectives, which do not guarantee such a latent space. We show that this leads to dark areas: regions of latent space that are reachable during diffusion sampling but decode to disconnected or chemically invalid molecules. Unlike in image generation, molecular decoding requires strict structural and chemical precision, so even small latent perturbations can produce catastrophic failures. We therefore propose TopVAE, a topology-optimized VAE that reduces dark areas by making the decoder internalize structural and chemical constraints during training, eliminating the need for test-time chemical correction. TopVAE greatly improves off-posterior robustness, and when paired with a standard DiT, achieves $77\%$ lower FCD-3D on QM9, the highest V&C, $52\%$ lower FCD-3D on GEOM-Drugs, and $1.29{\times}$ more stable and connected molecules on zero-shot scaffold inpainting.

We introduce c-Lattice Aggregation, a fault-tolerant reconstruction problem for distributed verification under crash and Byzantine failures. In our setting, n asynchronous processes supervise a concurrent execution I: each process holds a local sample, and must collaboratively reconstruct I from partial, potentially overlapping observations. A protocol solves c-Lattice Aggregation if at least c correct processes output the complete execution I, while all correct outputs are comparable and bounded by I. This strengthens Lattice Agreement [Attiya, Herlihy and Rachman, 1995] and Byzantine Lattice Agreement [Di Luna et al., 2020; Zheng and Garg, 2020]. We parameterize inputs by a redundancy parameter x -- every element of I appears in at least x initial samples -- and establish tight feasibility thresholds. Under crash failures with at most t faulty processes, Lattice Aggregation is solvable if and only if x >= t + 1. Under Byzantine failures with t < n/3, c-Lattice Aggregation is solvable if and only if x >= 2t + c. All bounds are tight: we present matching algorithms based on SCD-broadcast [Imbs et al., 2018; Khanchandani and Wattenhofer, 2024] and indistinguishability-based lower bounds. Finally, we define globally dependent languages -- those for which no partial view can certify correctness, including consensus, linearizability, k-set agreement, and leader election -- and prove that soundness of any monitoring system is achievable if and only if c-Lattice Aggregation is solved, yielding the first complete characterization of fault-tolerant verification under Byzantine failures.

Active Motor Noise Cancellation (AMNC) ships in commercial fused deposition modeling (FDM) 3D printers as a hardware countermeasure against acoustic side-channel attacks that target intellectual property (IP). We present the first empirical evaluation of a deployed AMNC countermeasure, using a public dataset of synchronized acoustic and vibration recordings from two AMNC-equipped Bambu Lab printers across 12 object classes. AMNC fully neutralizes the acoustic channel: classification accuracy is indistinguishable from the 8.33% random baseline. The vibration channel, which AMNC does not target, still leaks. With summary statistics the leak is coarse and amplitude-driven (vibration accuracy approximately 31% pooled, 36-47% within-printer), while the waveform shape carries essentially nothing (frequency-only features at chance). A full-sequence temporal model that ingests the ordered evolution of the print raises accuracy to approximately 61%, and an order-shuffling control (approximately 33%) shows that a substantial component is genuinely sequential and tied to print progression. The leak is device-specific: a classifier trained on one printer transfers near chance to the other. We conclude that AMNC is an acoustic-only defense: vibration remains a partial, geometry-correlated side channel it does not address, but one that does not, on this dataset, support full geometric reconstruction; reconstruction-grade attacks would require the magnetic or power channels AMNC also leaves untouched. We release all code.

Modern LLM-powered autonomous agents increasingly rely on rich user interface (UI) state observations to achieve reliable action grounding in complex digital environments. However, many deployments transmit the full UI state to remote inference servers even when most elements are irrelevant to the current task, which can leak sensitive but unnecessary context such as authentication codes, private notifications, and background application states. We propose MINIM, a trusted local broker that performs privacy-aware minimization on the client side before any observation leaves the device. Grounded in Contextual Integrity (CI), MINIM learns a dual-score representation for each UI element by predicting an inherent sensitivity score (s) and a task-conditioned necessity score (n). These scores drive a ternary disclosure policy that keeps essential elements, abstracts sensitive attributes when needed, and removes task-irrelevant content. We optimize a CI-aware objective that penalizes necessity errors more strongly on high-risk content, enabling aggressive pruning while preserving task-critical information. Experiments on real-world UI observations derived from WebArena show that MINIM substantially reduces task-irrelevant sensitive leakage while preserving task-critical semantic context and the interactive affordances required for reliable agent actions.

Rare diseases affect over $300$ million patients across more than $7{,}000$ conditions, yet no single hospital encounters enough cases of any one condition for reliable diagnosis. Cross-hospital collaboration could help by allowing a diagnosing institution to use distributed, case-specific diagnostic evidence, but privacy regulations restrict the transmission of identifiable clinical text across institutional boundaries. This setting raises two challenges: existing medical agent systems often rely on textual evidence exchange, while raw latent states such as hidden states and KV caches may still reveal prompt-derived clinical content. We introduce MedLatentDx, a latent multi-agent communication framework in which hospital agents keep private clinical records and retrieved cases local, and send compact latent KV blocks to a host agent for rare-disease diagnosis. MedLatentDx supports two deployment settings: same-backbone hospital agents use latent KV distillation, while hospitals with different LLM backbones use cross-family latent alignment. On CrossRare-Bench, a self-built large-scale rare-disease benchmark with hospital-level partitions, MedLatentDx improves cross-hospital diagnostic performance while reducing reconstructable clinical content relative to raw-latent communication baselines.

Large language models (LLMs) are increasingly characterised in recent evaluation work as having stable, model-level preference and value systems. However, accompanying robustness checks are limited to incidental prompt perturbations such as syntax variation and option reordering. This leaves open whether the measured properties survive when the surrounding task context changes, as it does in most real deployments. We test this directly across two established pairwise paradigms: ranking country preferences and eliciting utility judgements. In both, we make the deployment context -- the high-level task the model is performing while making concrete value-dependent choices -- our controlled variable, varied across framings such as writing a Reddit post or a news article. Across five LLMs and over 1.2M pairwise decisions, deployment context produces variation far larger than prompt paraphrasing and temperature controls. In country preference rankings over 15 countries, context induces widespread, statistically significant rank shifts; the aggregate Global North favouritism reported in prior work is itself context-dependent, with each model's bias shifting systematically across contexts. In utility elicitation over 50 outcomes, broad cross-category ordering is preserved, but fine-grained rankings within domains vary substantially, and cardinal exchange rates between outcomes (e.g. how many lives in one region equal one in another) shift by a factor of 2.47 at the median. Reported model-level preferences and utilities are therefore better understood as context-conditioned measurements than fixed model-level properties: safety guarantees obtained under one framing provide limited assurance in another.

The detection of gravitational waves has revolutionized our ability to explore fundamental aspects of the Universe. Traditionally, modeled gravitational-wave signals have been identified using template-based matched filtering, followed by coincidence analysis across multiple detectors in the signal-to-noise ratio time series. Recent advances in Machine Learning and Deep Learning have sparked growing interest in their application to both signal detection and parameter estimation. In this study, a hybrid Deep Learning strategy is proposed that leverages the effectiveness of Transformer encoders alongside well-established Convolutional Neural Network architectures in an attempt to estimate the intrinsic and extrinsic parameters of non-precessing binary black hole systems. The primary focus of this work is point estimation, producing single best-fit values for each parameter rather than full posterior distributions. This method is evaluated on both simulated signals embedded in Gaussian noise and real gravitational-wave events, and it demonstrates strong predictive performance and robustness across key astrophysical parameters.

Automated International Classification of Diseases (ICD) coding is a core medical-coding task for billing, epidemiology, and clinical decision support. Generative large language models (LLMs) are often reported as weak medical coders, but this finding mainly comes from inference-time settings such as prompting, retrieval, reranking, or tool use, leaving the role of task-specific post-training underexplored. We present a controlled empirical study of post-training for generative ICD coding, comparing discriminative baselines with LLM coders across prompting, supervised fine-tuning, and reinforcement learning under a common protocol and metric set. To our knowledge, this is the first study to evaluate RL-based post-training for generative LLM coders in ICD coding. We further introduce PHI, a diagnostic curriculum that extends GRPO to refine missed-code cases. Our results show that prompting-only evaluation substantially underestimates the potential of LLMs for ICD coding. SFT provides the main capability jump, GRPO further improves code-set prediction beyond SFT, and PHI provides targeted gains on macro-level performance. These findings suggest that the main bottleneck is not the generative formulation alone, but how the model is adapted and optimized for full-taxonomy recall. We release our code, data splits, and checkpoints at https://github.com/AlexandreWANG915/LLM4ICD.

Humans cannot always intuit what scenarios are most challenging to LLMs. Hoping to capture challenging edge cases, developers either design problems to be difficult for humans or curate extensive benchmarks. What if we could instead anticipate which scenarios a model will fail on? In this paper, we use an LLM's representational geometry to predict which concept combinations it will fail on. We attribute this compositional failure to interference between salient features. In tasks that require systematic composition - toy programmatic settings, multihop reasoning, multilingual factual recall - we find that when a pair of concepts is encoded near-orthogonally, the model reliably composes them. When their linear encodings are close, producing interference, the model fails to compose them. Our method reliably anticipates failure modes across different compositional tasks, without evaluating specific inputs. These results lay the groundwork to use representational geometry to identify high-risk examples, construct targeted stress tests, and provide a scalable foundation for active learning in real-world deployment.

Lawyer-client consultation is a critical starting point for legal services. Effective legal assistance hinges on eliciting sufficient and truthful information from clients in order to devise strategies that best protect their interests. This task requires Large Language Models (LLMs) not only to perform robust legal reasoning, but also to strategically elicit material facts through multi-turn interactions and effectively guide clients with diverse personalities. Yet existing legal benchmarks overlook this interactive capability. To fill this gap, we introduce DLawBench, a diagnostic benchmark for real-world legal consultation. Drawing on realistic client behavior, we characterize lawyer-client interactions into four types: Cooperative, Dependent, Withdrawn, and Adversarial. Using dialogues grounded in real cases, DLawBench evaluates whether LLMs can effectively conduct legal consultation under realistic conditions. DLawBench comprises 461 cases from Chinese and U.S. law, 5,532 paired fact entries, 3,411 inquiry rubrics, and 3,348 issue-resolution rubrics, and evaluates 26 representative LLMs. Systematic experiments show substantial headroom: the best-performing model, GPT-5.5, achieves only 0.562 on consultation-grounded legal reasoning. More importantly, DLawBench exposes both sycophancy in legal consultation and a paradox: models perform worse when clients need guidance most.

Vision-language models (VLMs) are typically trained as passive answerers, while their ability to actively ask diverse, non-trivial, visual-centric and grounded questions remains underexplored. Existing visual questioners' performance is bottlenecked by the availability of high-quality training data or the cost of curating them. We show that a VLM can continuously improve itself as a visual questioner without any external supervision. We propose a self-evolving framework that uses a VLM itself as both a proposer and a filter to produce harder, more informative, and visual-centric questions, while maintaining their exploration diversity to avoid training collapse. These questions are then used to train the VLM in both questioner and answerer modes. To evaluate the questioner, we introduce an agentic protocol that assesses questions along perception, reasoning, and diversity dimensions. Experiments across various backbone VLMs show that our method substantially enhances the quality and substantially expands the difficulty boundary of autonomous question generation. Under the same budget, our self-supervision is more effective than training on the static source data. Moreover, the self-evolving questioner remains a competitive or even better answerer.

Large language models can often close proof gaps in interactive theorem provers, but a verified theorem is not the same thing as a reusable library contribution. We study this distinction through a detailed case study: a semi-autonomous formalization of Grothendieck's vanishing theorem. The initial version compiles with no sorries, but an expert review found serious problems in definitions, theorem generality, file organization, and the API. We then ran a review-driven refactor and compression process and obtained a second expert review. The before-and-after comparison shows a sharp split: agents adapted well to local, mechanically checkable feedback, but remained weak at choosing definitions and designing APIs. We argue that autoformalization should be evaluated not only by closed sorries, but by whether the resulting formalization survives expert review.

Alzheimer's Disease (AD) is a progressive neurodegenerative disorder that affects millions of older adults, with prevalence expected to rise significantly in the coming years. Early diagnosis, particularly during the mild cognitive impairment (MCI) stage, is critical for timely intervention. Structural Magnetic Resonance Imaging (sMRI) has emerged as a key modality for detecting AD-related brain changes, but traditional graph-based approaches often struggle with modality and inter-site heterogeneity, limiting diagnostic performance. In this paper, we propose Graph Matching Network for Alzheimer's Disease Diagnosis (GMN4AD), designed to model interactions between heterogeneous brain graphs derived from neuroimaging data. Unlike conventional methods that treat each brain graph independently, GMN4AD leverages graph matching to capture cross-graph relationships, enhancing diagnostic precision. Furthermore, we introduce a test-time domain adaptation strategy that combines contrastive learning to mitigate domain shifts during inference. Extensive experiments on three public AD datasets demonstrate that GMN4AD achieves superior performance compared to state-of-the-art methods, offering a robust and generalizable solution for AD diagnosis.

We present a Bayesian calibration layer for slopsquat detectors -- those that flag hallucinated package imports in code produced by large language models (LLMs). Where existing pipelines emit binary decisions (flag / do-not-flag), our layer emits a Beta-posterior probability per detection, derived from a 3-category epistemic taxonomy that explicitly classifies each prior as empirically calibrated, constructively argued, or engineering-judgement-traced. Beyond the primary 200/404 registry channel, the calibrated layer exploits PyPI metadata signals -- package age, release count, author descriptor, summary -- to surface registered-but-suspicious packages that a binary registry detector misses, which is the realistic post-LLM-emission attacker regime. The resulting risk-aware primitive is directly consumable by downstream CI gates and supports principled threshold decisions across detection rules. We evaluate the calibration on a merged corpus of 1,734 Python snippets -- a stratified 189-prompt BigCodeBench slice plus a 100-prompt niche-library stress-test set, generated across a six-model panel spanning four cloud models (Claude-Sonnet-4.6, Mistral-Large, DeepSeek-v4-pro, DeepSeek-R1) and two local open-weight code models (Mistral Codestral, Meta CodeLlama). Against a re-implemented binary baseline inspired by Mahmud et al. -- which shares its registry oracle with our ground truth and therefore serves as a degenerate upper bound rather than a genuine competitor -- the calibrated layer reproduces the strict-registry detections and introduces well-calibrated additional flags on the metadata channel. We assess detector asymmetry with a McNemar paired test and calibration with both a flagged-subset Expected Calibration Error and a strictly proper full-corpus Brier score.

Each year, college admissions offices face an overwhelming challenge: processing millions of high school transcripts, each with unique formats, grading systems, and layouts. This manual process creates operational bottlenecks that delay admissions decisions and consume valuable resources. We present a transformative solution through a multi-agent AI system where specialized agents collaborate to automatically process diverse transcript formats through intelligent coordination and communication. Our multi-agent architecture consists of three specialized agents-a Pattern Recognition Agent for format-specific parsing, a Semantic Analysis Agent for natural language understanding, and a Vision Intelligence Agent for multimodal document analysis-coordinated by an Orchestration Agent that manages agent communication and result reconciliation. Our key innovation lies in agent-based quality control using GPA extraction as a coordination signal, ensuring reliable agent collaboration and preventing critical information loss. When evaluated on 40 real world transcripts from high schools across 13 U.S. states, our agent system successfully processed every document, achieving 96.7% accuracy compared to expert manual review while maintaining practical processing speeds of 45 seconds per transcript. This work demonstrates how multi-agent coordination can solve complex document processing challenges, offering institutions a scalable, collaborative AI solution that preserves accuracy while dramatically reducing processing time.

Neural-network quantum states (NQS) are a leading variational tool for quantum many-body physics, yet their optimization is fragile whenever the ground state carries a non-trivial sign or complex phase structure, a situation generic to gauge fields, broken time-reversal symmetry, and fermionic statistics. We trace this fragility to the stochastic estimator of the phase gradient rather than to network expressiveness. The phase sector of the Monte Carlo energy gradient is a noisy score-function estimator; differentiating the local energy instead yields a direct estimator that is unbiased for the same phase force, has far lower variance, and requires only a separated amplitude--phase ansatz. Demonstrated on a 100-site flux ladder, a small network trained this way reaches $0.89\%$ median error, where tuned standard baselines plateau at $1.8\%$ and wider or deeper standard-gradient networks degrade from $8.4\%$ to $24.6\%$. The advantage carries over to chiral XXX chains: the direct estimator again converges to a markedly lower error than the standard one, across $α$ and size; it grows with flux and vanishes in zero-flux controls. An adaptive-mixture of the two estimators is provably never worse in variance than the better endpoint at the optimal mixing coefficient, with seed-resolved diagnostics tracing much of the gain to eliminating failed runs. Estimator design thus emerges as a first-class lever for complex-valued neural quantum states.

LLM-based query expansion improves retrieval by enriching the original query with additional context. Yet most methods remain generation-driven, producing plausible pseudo-documents or expansions without checking how the target corpus responds. This can introduce retrieval drift, amplify misleading vocabulary, or miss terms that distinguish relevant from non-relevant documents. We argue that effective expansion requires retrieval-grounded feedback, not just single-pass generation or unverified iteration. We introduce ADORE (ADapt, Observe, Relevance Evaluate), an iterative framework that turns retrieval outcomes into feedback for the next expansion. At each round, an LLM generates pseudo-passages, a retriever exposes the corpus response, and a relevance assessor evaluates retrieved documents against the original query. These judgments identify what to reinforce, what remains undercovered, and what to suppress. Across TREC Deep Learning, BEIR, and BRIGHT, ADORE consistently outperforms strong query expansion baselines with notable improvements across nearly all evaluation settings, improving average nDCG@10 by 24.5% over BM25 and 3.6% over the strongest prior query expansion method on BEIR, and by 122.9% over BM25 and 9.2% over the best query expansion baseline on BRIGHT. Our code and data are publicly available.

Exploratory question answering (EQA) over data lakes requires an LLM agent to discover relevant sources, analyze retrieved data, and adapt its actions based on intermediate results. End-to-end accuracy alone cannot distinguish failures in search, planning, data analysis, or the agent's Action Policy: its decisions about what to do next and when to submit an answer. We present SANA (Search Agent Navigation Ablation framework), a diagnostic ablation framework that transforms EQA tasks into runtime profiles containing gold source sequence, sanitized subquestions, and execution records. SANA uses these profiles to construct idealized search, planning, and data-analysis tools, allowing each component to be ablated; the residual gap is diagnostic evidence for policy failures. To illustrate SANA as a reusable evaluation framework, we adapted two recent EQA benchmarks, LakeQA and KramaBench, and evaluated lightweight and mid-sized agents under fixed prompts, budgets, data lakes, and runtimes. Across both benchmarks, data analysis is a consistent bottleneck while planning is less so. Search is a major limitation in LakeQA's large data-lake setting, but less so for the smaller-scale KramaBench. SANA thus deconstructs end-to-end task accuracies into a diagnosis of where data-lake agents fail, and allows for systematic comparisons of progress in search, planning, data analysis, and agent design.

Spiking Neural Networks (SNNs) have emerged as an energy-efficient alternative to conventional neural networks, demonstrating strong performance in computer vision and robotics. More recently, SNNs have been applied to time series forecasting (TSF), with methods exploring spiking temporal backbones, spike-compatible positional encodings, Fourier-domain processing, and redesigned neuron dynamics. However, existing SNN forecasting approaches process variables independently, lacking explicit mechanisms for modeling inter-variable dependencies. This is a critical limitation in multivariate settings, where cross-variable correlations carry substantial predictive information. We propose Spiking Fourier Graph Operators (SpikF-GO), which addresses this gap by combining a hypervariate graph formulation in which every scalar observation becomes a graph node with spike-driven spectral processing. SpikF-GO introduces a Hard Concrete frequency gate for learnable sparse frequency selection and a Complex LIF gate that applies independent spiking neurons to real and imaginary Fourier components, preserving binary, event-driven computation throughout the spectral domain. We further present a variant incorporating Central Pattern Generator-based positional encodings for stronger long-range temporal modeling. Evaluated on eight benchmarks under a unified experimental protocol, SpikF-GO achieves the best average rank among all SNN methods and outperforms its ANN counterpart, FourierGNN, at reduced energy cost. SpikF-GO maintains competitive accuracy even at substantially smaller embedding dimensions, thereby achieving significant energy reductions. To our knowledge, this is among the first works to bring graph-based multivariate modeling into the spiking domain for TSF and the first to provide a unified comparison across SNN forecasting architectures under a common experimental protocol.

Creative image editing tools, such as Photoshop's Remove or Generative Fill buttons, are central to everyday customer use and account for a major share of traffic in Photoshop and Lightroom. However, current generative AI models face significant latency challenges, which become even more pronounced when transitioning from convolution-based U-Nets to Diffusion Transformers (DiTs). In our evaluation on hundreds of representative image editing samples spanning a wide range of mask ratios, the DiT module alone accounts for an average of 73% of the total model latency, even after being distilled from 50 timesteps down to 8 timesteps. To tackle this challenge, we propose $\textbf{HiLo-Token}$, an input-adaptive token compression framework that allocates more token budget to high-frequency, rich-context regions while assigning fewer tokens to low-frequency areas. Specifically, for the editing region specified by the user mask, we retain all tokens within a dilated mask to preserve strong locality and contextual relevance. Outside the editing region, we introduce a simple yet effective high-frequency token selection strategy based on spatial frequency to capture important local details, while using tokens from a 16x downsampled image to represent low-frequency components and preserve the blurry but global structure. Extensive experiments on production-level evaluation data validate the effectiveness of the proposed method, achieving 3.13x, 2.59x, and 1.67x DiT speedups on A100-80GB for image editing tasks across small, medium, and large mask ratio categories with average ratios of 6.38%, 15.92%, and 35.36%, respectively, without any regression in generation quality.

Self-supervised geospatial foundation models (GeoFMs) learn transferable representations from remote sensing data, but their downstream behavior is difficult to characterize. We study six representative GeoFMs spanning joint-embedding, reconstruction, and multimodal pretraining families, and evaluate transfer across classification, regression, and segmentation benchmarks under different label availability and downstream pipelines. We find that model rankings change across tasks and adaptation settings. Layerwise probing shows that, in most cases, task-relevant information is more accessible in intermediate transformer blocks compared to final-layer embeddings, and that GeoFMs exhibit distinct depthwise profiles. In segmentation case studies on PASTIS and Sen1Floods11, downstream adaptation settings such as decoder design and fine-tuning can be as impactful as the choice of GeoFM, and standard dense-prediction heads may be poorly aligned with how GeoFMs organize information over depth. Finally, CKA analysis on case studies shows that fine-tuning does not rewrite GeoFMs uniformly across depth, and the strongest changes are localized to the first linear layer of the MLP in ViT blocks. These results help explain why GeoFM rankings shift across benchmarks and motivate more representation-aware evaluation and adaptation strategies.

AdamW is a default optimizer for modern deep learning, but its first and second moment states add roughly two parameter-sized buffers to training memory. We propose Gefen, a memory-efficient optimizer that automatically shares second-moment estimates across parameter blocks and quantizes the first moment using a learned codebook, thereby reducing AdamW's memory footprint by ~8x while maintaining the same performance, corresponding to a reduction of 6.5 GiB per billion parameters. The method is motivated by a theoretical result showing that large mixed Hessian entries constrain the ratio of squared gradients toward one, suggesting that Hessian-aligned parameters are natural candidates for sharing second-moment statistics. Since computing Hessians is impractical at scale, Gefen infers block structure from the initial squared gradients, requiring no architecture-specific metadata or hyperparameters beyond AdamW defaults. Gefen learns an exact histogram-based dynamic-programming quantization codebook and reuses the same blocks for first-moment scaling. Across diverse experiments, Gefen achieves the lowest peak optimizer memory among the compared AdamW-like methods while maintaining AdamW-level performance. In FSDP and DDP training, the reduced memory footprint enables larger microbatches and improves throughput significantly over AdamW, providing a practical drop-in replacement with lower memory usage that can increase throughput and enable training larger models or using larger batch sizes. We provide the complete Python implementation, including fused CUDA kernels at https://github.com/ndvbd/Gefen

The intersection of crypto x AI is spawning papers, products, online posts, and companies. All the surrounding buzz, though, obscures what exactly has been done, what the opportunities and challenges are, and what open questions deserve attention. This survey paper asks what AI can do for blockchain-based technologies (broadly construed as "crypto") (crypto x AI), and vice versa (AI x crypto). We systematize existing work, summarize key takeaways, highlight open research questions, and offer a perspective on pervasive industry misconceptions, concluding that AI and crypto are still in the very early stages of meaningful integration.

Vision-Language-Action (VLA) models excel at mapping visual inputs and natural language instructions directly to robotic control policies. However, because they are trained primarily to fit behavioural demonstration data, they do not explicitly enforce fundamental physical principles such as rigid-body dynamics or contact constraints. This exposes a critical physics gap: standard temporal smoothing applied on top of single-step or chunked VLAs trades trajectory quality for added failures that short-term memory cannot resolve. To bridge this gap, we introduce PhysVLA (Physics-VLA), a plug-and-play, inference-time framework designed to wrap any frozen VLA backbone without retraining, fine-tuning, or weight access, with less than 1 ms of overhead per control step. PhysVLA intercepts the predicted control action, captures only the simulator or system state, and applies a dual-layered correction: (i) a phase-aware finite-state machine that structures discrete task segments (approach, grasp, transport, and place), and (ii) a selective Euler-Lagrange gate that activates only when a dynamics oracle detects kinodynamic inconsistency. Evaluated across OpenVLA, OpenVLA-OFT, Force-VLA, and Generalist-VLA on LIBERO-Spatial with a 7-DoF Franka Panda, the framework delivers absolute success rate increases of up to 17% and stability increases of up to 19% with no per-task regressions, improves trajectory efficiency by up to 15% across all four backbones, and shows up to a 10x improvement in trajectory jerk robustness on a Robosuite Lift cross-simulator sweep. We further validate the framework on a real Agilex Piper arm with a pick-and-place task, confirming that PhysVLA transfers to physical hardware without retraining, with success-rate improvements of up to 50%, establishing physical awareness as a composable, backbone-agnostic runtime module.

Modern decision systems increasingly rely on learned components whose outputs may be confident yet wrong, exposing downstream actions to costly errors. We introduce Risk-Aware Causal Gating (RACG), a framework that decides whether to act on, defer, or abstain from a model's prediction by combining causal effect estimation with calibrated risk control. RACG models the causal pathway from candidate actions to outcomes and gates each decision according to an estimated counterfactual risk rather than raw predictive confidence. To make gating reliable, we derive distribution-free bounds on the probability of acting under high-risk conditions and show how these bounds translate into operating thresholds that satisfy user-specified safety constraints. We further propose an adaptive gating policy that adjusts to distribution shift by monitoring discrepancies between predicted and realized outcomes, tightening the gate when causal assumptions appear violated. Across simulated interventions and real-world decision benchmarks, RACG reduces high-cost errors substantially while preserving most of the utility of an ungated policy, and it outperforms confidence-based and selective-prediction baselines at matched abstention rates. Our results indicate that explicitly separating causal risk from predictive uncertainty yields decision systems that are both safer and more transparent, offering a principled mechanism for trustworthy automation in high-stakes settings.

Background: AID algorithms require ongoing software updates and bug fixes. In co-adapted systems, where users tune settings around existing algorithmic behavior, bug fixes can paradoxically disrupt glycemic control. No principled framework evaluates the safety of AID algorithm updates. Methods: Our two-part framework classifies bugs and evaluates the clinical equivalence of AID system software updates. Bugs are classified as factual, heuristic, or computational, each with distinct management strategies. Classifications were validated from porting Trio's oref algorithm from Javascript to a bug-fixed Swift implementation. We compared implementations using shadow execution on 736,480 invocations from eight Trio users. The second component assesses clinical equivalence with error analysis on paired glucose values, applied to both Trio implementations using mechanistic in silico and data-driven replay simulation. Results: In mechanistic in silico simulation, the Swift and Javascript implementations produced nearly identical Time in Range (84.9% vs. 84.9%) and Glycemia Risk Index (23.5% vs. 23.9%), with more than 99% of paired glucose in Parkes Error Grid Zones A and B, meeting our clinical equivalence threshold. Shadow execution showed low mismatch rates in oref components (iob 0.43%, autosens 1.22%, determineBasal 0.07%, meal 0.01%), with clinically meaningful differences in 0.03% of iob invocations. Data-driven replay simulations of bugs revealed more than 99% of downstream paired glucose in Parkes Error Grid Zones A and B, also meeting our clinical equivalence threshold. Conclusions: Our framework integrates bug-fixing principles with multi-method clinical evaluation to assess AID algorithm update safety. It is system-agnostic and applicable to all widely used OS-AID systems, with case studies highlighting the need for systematic remediation of factual and computational bugs.

Accurate estimation of long-term care transition probabilities is central to disability insurance pricing, reserving, and solvency assessment. Classical actuarial multi-state models commonly rely on Markov, semi-Markov, or proportional-hazard specifications, which provide a direct connection to cohort projection but may be restrictive for irregular longitudinal health data with nonlinear aging patterns and heterogeneous covariate histories. This paper develops a well-calibrated estimator of multi-state transition probabilities for irregular longitudinal health data. The model learns from individual health history, incorporates the time elapsed between observations, and conditions transition probabilities on demographic and socioeconomic attributes. It produces a valid probability distribution over the next observed health state, with four possible states: healthy, mild disability, severe disability, and death. Individual probabilities are aggregated by age group and origin state to form transition matrices compatible with actuarial cohort projection. Using longitudinal data from the Health and Retirement Study, we compare the proposed estimator with logistic regression, gradient-boosted trees, a recurrent neural network, and a last-state persistence benchmark. The evaluation considers probabilistic accuracy, endpoint discrimination and calibration for severe disability and death, risk concentration, and transition matrix error after aggregation. The proposed estimator improves severe disability discrimination relative to logistic regression and gradient-boosted tree benchmarks, maintains strong calibration, and yields the lowest transition matrix error among the evaluated models in the held-out test analysis. Results show that a structured machine learning estimator can support long-term care transition modeling when judged by calibration and projection fidelity, beyond discrimination.

Unlearning aims to remove the influence of specific training data sources, but this has proved challenging because the contributions of different sources are entangled within the model. Isolating source contributions to disjoint parameters makes removal easier, though it obstructs joint learning across sources. We propose NULLs (Natively Unlearnable LLMs), a model class that satisfies the two opposing goals of isolating source-specific contributions and learning jointly across sources, by training a set of shared backbone neurons alongside a pool of sparsely activated sinks. During training, information specific to a source naturally concentrates in its sinks while information shared across sources accumulates in the backbone. A source is then unlearned at deployment by disabling its corresponding sinks, with no gradient updates and no access to the retained data. We show that NULLs scales to Wikipedia's ~6M articles, isolating each as an independent source. Unlearning a single article removes knowledge specific to it while preserving facts shared with semantically related articles, closely matching retraining from scratch. We note that unlearning with NULLs is also robust: in a case study of unlearning the Harry Potter books, NULLs resists both adversarial extraction and relearning that reverses post-hoc unlearning. Finally, NULLs preserves general language capabilities, matching a standard transformer on downstream benchmarks. Together, these results suggest that source-level unlearning need not be an afterthought. It can be built natively into LLM training while retaining the benefits of shared representation learning.

Tabular data embeddings have become a cornerstone of data profiling and data integration pipelines, enabling tasks such as entity annotation and resolution; schema matching; column type detection; and table search, among others. Existing approaches embed rows, columns, or entire tables into a vector space and rely on nearest-neighbor search to retrieve candidate matches. A fundamental limitation of current embedding methods is the lack of interpretable similarity scores: the concrete similarity value between a query and its nearest neighbour carries no intrinsic meaning, making it impossible to determine whether that neighbour is a true match or simply the least-dissimilar item in a corpus that contains no valid answer. This inability to set principled thresholds for retrieval undermines practical deployment, particularly for zero-match detection. We investigate the use of HyperDimensional Computing (HDC), specifically the Holographic Reduced Representations (HRR) model, as a framework for tabular row embeddings when the retrieval task corresponds to answering structured select-project queries in vector space. Exploiting the algebraic properties of HDC operations, we derive closed-form expected similarity values for both equality and non-equality retrieval predicates, which converge to interpretable values as dimensionality increases, and use these to identify suitable retrieval thresholds. We evaluate HDC against EmbDI, a graph-based baseline, on two real-world datasets across varying table sizes and predicate lengths. Our results show that HDC matches or outperforms EmbDI for row retrieval across all configurations, handles non-equality predicates more robustly, and achieves perfect attribute projection accuracy at sufficient dimensionality -- while uniquely enabling reliable identification of zero-match predicates through its principled thresholds.

Vision-language models (VLMs) can answer image-based questions confidently, and often correctly, even when no image is provided. This mirage behavior inflates benchmark scores without reflecting visual grounding. Prior work treats this as a single failure mode. We argue it is two. Using Mirage Probes, a contrastive probing framework that pairs paraphrased question variants with matched mirage and non-mirage labels on the same image, we show that mirage behavior is linearly decodable from internal activations across residual stream, MLP, post-attention, and attention-head sites in two open-source VLMs. We demonstrate that a Naive Bayes text baseline cannot recover this signal, ruling out surface lexical confounds. Cross-benchmark separability patterns, together with a novel Prior Harnessing Index (PHI) measuring how much a model can answer from text alone, expose two distinct regimes: textual biases, where the model answers from language priors without engaging visual representations, and spurious images, where it constructs false visual content in latent space and answers as if grounded. The distinction has direct mitigation consequences: text-distribution cleaning can address the first regime but cannot reach the second, since spurious-image mirages live in the model's visual representations rather than its text. Faithful visual grounding will require interventions at the representational level.

We present an end-to-end pipeline for estimating stellar parameters from Sloan Digital Sky Survey Data Release 12 spectra using a fully connected multitask neural network with residual blocks, whose hyperparameters are tuned via Bayesian optimization. The preprocessing pipeline includes per-spectrum standardization, RobustScaler normalization of the target variables -- effective temperature $T_{\mathrm{eff}}$, metallicity $[\mathrm{Fe/H}]$, and surface gravity $\log g$ -- and data augmentation via Gaussian noise injection. On a held-out test set, the model achieved Mean Absolute Errors (MAE) of $59.76~\mathrm{K}$ for $T_{\mathrm{eff}}$, $0.103~\mathrm{dex}$ for $[\mathrm{Fe/H}]$, and $0.130~\mathrm{dex}$ for $\log g$. Normalized against the full-scale range of each parameter, these results represent range-normalized errors between $1\%$ and $3\%$, achieved with a highly efficient model complexity of approximately 540,000 trainable parameters. These results demonstrate that a compact residual multitask architecture, combined with principled signal preprocessing, provides a parameter-efficient solution for nonlinear parameter estimation in large-scale spectral datasets. In particular, the proposed model achieves competitive performance with substantially lower complexity than deeper neural network baselines.

Muon is an increasingly widely used optimizer that replaces a gradient $G=USV^\top$ with its polar factor $UV^\top$, thereby flattening the singular spectrum. However, full flattening discards singular-value information that may matter for adaptation. We introduce Muon$^p$, a Muon-style optimizer that instead uses fractional spectral-power updates $US^pV^\top$ for rational $p\in(0,1)$, interpolating between Muon and gradient descent. To make it practical, we prove that fractional spectral powers cannot be computed by any fixed univariate polynomial iteration, and furthermore derive low-degree odd bivariate recurrences that approximate $US^pV^\top$ using only matrix multiplications, preserving Muon's matrix-multiplication-only structure and compute complexity. We show that Muon$^p$ maximizes the linear improvement in loss under the Schatten $q$-norm for $q=1+\frac{1}{p}$. Empirically, Muon$^p$ is especially effective for finetuning: on billion-scale models, Muon$^p$ improves validation perplexity and downstream task performance. We further analyze when Muon$^p$ is less suitable, through the lens of spectral geometry. Our results reveal important insights on when preserving the singular spectrum can bring significant gains, and introduce a principled way to achieve them.

Hardware Security at Willamette is a Willamette University affiliated research group studying the hardware-software interface of security critical services. Within our program, we noticed many researchers spent considerable development time learning to understand and manually parse traces-of-execution of hardware designs which are used to identifying whether vulnerabilities or weaknesses arise at the hardware, software, or interface level. We propose the "RTL-Arrow" framework, a framework to compile performant binaries which bridge the hardware/data divide. We translate the outputs of simulated hardware execution, as "value change dumps" into modern data science workflows as cloud-ready "dataframes", to standardize program verification across the hardware and software levels. We describe our approach, its benefits, and lessons learned from the process of packaging and distributing these libraries for our security research program.

Long Chain-of-Thought (CoT) reasoning improves LLM problem-solving but is computationally expensive due to sequential token generation. While recent works explore reasoning in continuous latent spaces to bypass discrete token generation, they often struggle with training stability and fail to scale to complex, long-horizon tasks due to lack of supervision signal. We propose SuperThoughts, which compresses pairs of consecutive CoT tokens into single latent representations and decodes two tokens per step via a lightweight Multi-Token Prediction (MTP) module. This preserves discrete token supervision at training time while doubling throughput at inference time. We finetune Qwen2.5-Math-1.5B-Instruct, Qwen2.5-Math-7B-Instruct, Qwen2.5-Math-14B-Instruct, and evaluate on MATH500, AMC, OlympiadBench, and GPQA-Diamond. With a confidence-based adaptive mechanism that falls back to standard decoding when uncertain, SuperThoughts achieves $\sim$20--30\% CoT length reduction while maintaining accuracy with minimal degradation (1-2 points accuracy drop on most tasks).

Security validation is an important yet challenging part of the hardware design process, yet, by convention, validation engineers are tasked with defining the threat model, specifying the relevant security properties, detecting any violations of those properties, and assessing the consequences to system security, each of which is manually intensive and may introduce errors. The combined technologies of information flow tracking and specification mining represent an automated approach to property generation and validation, but prior work on information flow tracking on RTL trace data was limited to find cases under which information flowed between registers, without reproducing full paths to capture how sensitive information propagates through a design. With the introduction of new technologies accelerating hardware analysis, we develop a novel approach for constructing information flow paths from register transfer level (RTL) trace data.

Many materials and chemical systems exhibit history-dependent responses, where functional outcomes are governed not only by final-state variables but by the time-dependent sequence of fields, temperatures, or chemical potentials applied during operation. Discovering new processing protocols is therefore a high-dimensional search problem in which the control variable is an entire waveform or sample history, and conventional strategies either remain confined to conservative interpolative families or become prohibitively measurement intensive. Here, a closed-loop workflow is introduced that couples evolutionary search over a compact waveform representation with uncertainty-aware deep kernel learning to generate, rank, and experimentally validate candidate protocols. Applied to ferroelectric thin films, with the scanning-probe tip-bias waveform as the protocol and the nonlinear electromechanical response as the reward, the workflow discovers waveform families that enhance nonlinearity by de-aging the film. Spatially resolved before/after measurements show that the best-performing waveforms selectively activate pre-existing, weakly pinned domain-wall segments, whereas the worst drive long-range irreversible switching. This framework reframes protocol tuning as out-of-distribution discovery, generalizable to synthesis and annealing trajectories, battery formation protocols, and other high-dimensional control problems.

Recommendation systems are essential in modern music streaming platforms due to the vast amount of available content. While collaborative filtering is widely used to suggest items based on the preferences of others with similar patterns, it performs poorly in domains where user-item interactions are sparse, such as music. Content-based filtering is an alternative approach that examines the qualities of the items themselves. Genre, instrumentation, and lyrics have been explored; however, relatively little attention has been given to emotion recognition. Since a user's emotional state strongly influences their music choice, incorporating mood signals offers a promising direction for personalization. In this work, we propose a mood-conditioned ranking framework that integrates user affective signals into the recommendation process via softmax-based sampling in the energy-valence space. We evaluate the approach via single-blind experiments in which participants compare recommendations from the proposed system against a baseline. The results indicate improved perceived recommendation quality, providing preliminary evidence for the effectiveness of incorporating mood-based inputs into music recommendations.

We present SpheriCity, an expert-grounded conversational prototype designed to support trustworthy knowledge sensemaking from sustainability reports. City-level circularity assessment reports contain rich information about materials, infrastructure, and policy interventions, yet their length and heterogeneous structure make cross-document synthesis and comparison difficult for practitioners and researchers working on circular economy initiatives. While large language models (LLM) promise faster knowledge access and synthesis, their opaque reasoning, hallucinations, and lack of source transparency introduce risks for trust and interpretability, and require verification in high-stakes sustainability contexts. SpheriCity addresses these challenges through a provenance-first conversational agent that foregrounds evidence traceability, structured synthesis, and interaction scaffolds to support exploratory querying and cross-document synthesis across sustainability reports. We conducted a formative expert review with six sustainability experts using representative queries spanning cross-city comparison, policy summarization, and recommendation-oriented tasks. Experts evaluated responses across dimensions and provided qualitative reflections on the system's usefulness for sustainability knowledge work. Our results reveal that transparent sourcing, contextual explanation, interpretability, and alignment with expert workflow strongly shape expert trust and judgments of system usefulness. This work contributes (1) a conversational prototype for sustainability knowledge sensemaking, (2) an expert-grounded evaluation framework for assessing AI responses in high-stakes knowledge domains, and (3) design insights into how provenance, uncertainty communication, and integration in workflow influence expert users' trust in AI assistance for sustainability decision support.

Neural topic models enable scalable semantic discovery, but their integration with quantum hardware remains largely unexplored. We present a proof-of-concept hybrid classical-quantum variational autoencoder (VAE) for topic modeling, embedding parameterized quantum circuits within the VAE inference network while retaining a classical topic-word decoder. To address the resource constraints of quantum hardware, we propose a modified Gaussian Softmax posterior that decouples latent space dimensionality from the number of topics to be extracted, enabling the model to operate with a low-resource 10-qubit quantum device. On the AgNews dataset, the hybrid VAE outperforms state-of-the-art neural topic models (NTMs), reaching a $C_v$ coherence score of 0.71 and an NPMI score of 0.20 while preserving high topic diversity. For comparison, we also construct a fully classical variant, which also outperforms state-of-the-art models on AgNews and exhibits clear class separation in the latent space. These results demonstrate that hybrid VAEs are computationally viable even on NISQ-era devices and represent a promising direction for quantum-enhanced topic modeling.

Mobile networks continue to grow in complexity and next generation networks are expected to support both increasing traffic loads and more diverse services. As network complexity rises, optimizing antenna parameters under dynamic or changing objectives becomes increasingly challenging. We propose a novel multi-agent reinforcement learning (MARL) algorithm for high-level control and orchestration of mobile networks. The Temporally Consistent Graph Q-Network (TC-GQN) algorithm learns a self-predicting representation of the whole network that is task-independent and aggregates information from all base-stations. A graph neural network is trained using a global reward function to assign coordinated local actions based on the learned encoding of the global network state. We evaluate the algorithm in a simulated environment to orchestrate an energy-saving feature across multiple sectors and multiple carriers under different quality of service (QoS) constraints. The proposed algorithm outperforms state-of-the-art graph-based baselines and a competitive rule-based controller by improving hardware sleep time while maintaining QoS. Moreover, the learned representation enables rapid adaptation to changing intents.

Ramulator 2.1 is a major overhaul of Ramulator 2.0 that substantially improves the simulator in three directions: 1) support of modern and emerging DRAM and memory-controller features, 2) better usability and extensibility of the simulator, and 3) more comprehensive tests and validation workflows. Ramulator 2.1 adds support for advanced features in recent and emerging DRAM standards and memory controllers, including HBM3/4, LPDDR5/6, and GDDR7. To improve usability and extensibility, Ramulator 2.1 introduces a Python-based modeling and configuration interface backed by a two-way code-generation framework that 1) hides low-level C++ code behind high-level DRAM specifications written in Python, and 2) automatically creates Python proxies for all components of the simulator. Doing so enables users to rapidly create variants of DRAM standards and automate design-space-exploration workflows. To improve trustworthiness in simulation results, Ramulator 2.1 provides a comprehensive testing and validation infrastructure that covers both 1) fine-grained validation of specific DRAM timing constraints and memory-controller scheduling behavior, and 2) system-level performance evaluation using latency-throughput curves. To aid performance analysis and debugging, Ramulator 2.1 also includes an easy-to-use and high-performance DRAM command trace visualizer. Ramulator 2.1 is open-source on GitHub and under active development.

Remote photoplethysmography (rPPG) transformers achieve low heart-rate error on benchmarks, yet their decisions remain opaque--a growing concern as rPPG moves toward clinical heart rate estimation. Existing rPPG XAI is dominated by qualitative heatmap inspection without quantitative faithfulness metrics or physiology-grounded validation, leaving a gap between visual plausibility and auditable evidence. We address this gap. First, we adapt four attribution methods (raw attention, rollout, flow, Beyond Intuition) to RhythmFormer's bi-level routing attention with top-$k$ selection. Second, we introduce a skin coverage metric quantifying how much attribution mass falls on skin regions. Third, we adapt the SaCo faithfulness coefficient from its original classification setting to rPPG regression by using the MAE between original and perturbed predicted rPPG waveforms as the perturbation impact. Applying these tools, we quantify a multi-hop leakage effect under sparse top-$k$ routing: attention rollout and flow almost completely restores the connections that individual refined-attention layers explicitly set to zero. Beyond Intuition mitigates this via its value-projection-weighted rollout and gradient-supported mask, attaining the highest median refined skin coverage ($0.83$ vs. $0.57$ for vanilla rollout) and faithfulness ($F=0.92$) among the evaluated methods on UBFC-rPPG. Validation across diverse datasets and model variants is needed. A case study on a low-SaCo outlier further shows all four methods recovering consistently once an artefactual region is replaced, suggesting consistent SaCo behavior across attribution families in this illustrative case. Together, these metrics move XAI for rPPG toward auditable numerical evidence about spatial alignment and perturbation faithfulness, i.e. trustworthy rPPG XAI.

LLM-based generative agents are increasingly used in urban simulators, yet it remains unclear whether they reproduce empirically realistic human mobility patterns or merely generate plausible mobility narratives. We introduce a validation framework for evaluating the mobility of generative agents of LLM-based urban simulators against real-world mobility data. For this, we use mobility laws, temporal rhythms, network motifs, semantic activity transitions, and behavioral mobility profiles. Using datasets from the Greater Paris region and Shanghai, we evaluate AgentSociety and CitySim across multiple dimensions of mobility realism. Our analysis reveals a substantial gap between narrative plausibility and empirical mobility realism. Although the simulators capture some high-level semantic activity distributions, they struggle to reproduce core spatial and temporal constraints, including realistic trip-length distributions, origin-destination flows, dwell times, and transition dynamics. We further observe that realistic mobility diversity is unstable across default prompting configurations and may require explicit profile-aware initialization. To support reproducible evaluation, we also contribute scalable and open LLM-driven infrastructure for regional-scale map generation, observability-enhanced simulation, mobility-metric computation, and traffic simulation. Our findings highlight the need for rigorous empirical validation of LLM-based urban simulators and provide practical tools for building more realistic and reproducible urban simulation systems.

Many NP-hard graph problems can be modeled as optimal subgraph extraction problems with feasibility constraints. From Network Design to Facility Location, from Robotics to Graph Drawing, the subgraph extraction pattern emerges across diverse domains. Despite this commonality, these problems are typically solved with domain-specific heuristics. Usually, these problems balance competing objectives such as maximizing coverage or minimizing cost while satisfying structural constraints such as connectivity, planarity and reachability. In this work, we introduce $Δ$Search, a general and fast heuristic framework that exploits the insight of Reward-Penalty optimization for solving a large class of subgraph extraction problems. The framework is easy to use as it only requires feasibility constraints and optimality criteria to be provided by the user to express the subgraph extraction problem. We also show how exact methods can be augmented with $Δ$Search to improve their performance by aggressive pruning of the search space. We evaluate our framework on monotone graph problems such as Maximum Planar Subgraph (MPS) and Minimum Connected Dominating Set, Weighted Monotone problems such as Maximum Weighted Independent Set and Minimum Weighted Steiner Tree, and non-monotone graph problems such as Prize Collecting Vertex Cover (PCVC) and Uncapacitated Facility Location Problem (UFLP). Our results show that $Δ$Search matches or surpasses state of the art heuristics for MPS, UFLP and PCVC problems with similar runtime. For the remaining problems, $Δ$Search achieves approximately 89% of the solution quality of the state-of-the-art algorithms without any problem-specific tuning

Autonomous network-security response systems promise to reduce Security Operations Centre (SOC) reaction latency, but reward-only multi-agent reinforcement learning (MARL) can improve security reward while remaining non-deployable. We present a safety-contract graph MARL framework and instantiate it as ACD$^3$-GAT (Adaptive Constrained Counterfactual Decisioning with a Graph Attention Network encoder), an architecture that separates simulator observations from reusable operational budgets, constrained optimization, graph state encoding, and counterfactual action screening. We evaluate the method in CAGE Challenge 4, where agents operate under budgets for Mean Time to Recover (MTTR), false-positive response, and firewall change-management disruption. Across the benchmark, every unconstrained method violates the SOC downtime budget in 100% of evaluated episodes, with mean downtime proxy costs of 311-430 against a budget of 50. This complements prior CAGE Challenge 4 findings by showing that reward-only learning lacks operational discipline. Constrained MAPPO-GAT (C-MAPPO-GAT) isolates Lagrangian operational-cost control and budget-aware screening, while ACD$^3$-GAT adds budget context, CVaR tail-risk estimation, opponent-belief state, and Graph Counterfactual Risk Propagation (G-CRP). The replicated comparison includes three 200-episode seeds for IPPO, MAPPO-GAT, C-MAPPO-GAT, and ACD$^3$-GAT. C-MAPPO-GAT reduces downtime violation from 100% to 0.3% and mean downtime cost from 355.4 to 15.5 relative to MAPPO-GAT. ACD$^3$-GAT reduces mean downtime cost to 48.2 with a 13.8% violation rate, placing it on the safety-contract frontier rather than at the most conservative compliance point. Topology-seed and coupled adaptive Red-process stress tests preserve this contrast and show lower worst adaptive degradation for safety-constrained policies than reward-only MAPPO-GAT.

We can do more than defend science from a flood of AI-assisted papers. Used well, AI offers a historic opportunity to correct distortions in the publication system, help us publish fewer and better papers, and give scientists back the time to do their best work.

Markovian Whittle-Matérn fields have been convergently approximated by discrete Gauss Markov Random Fields (GMRFs) with sparse precision matrices using a Finite Element approximation of the two-parameter family, \[ (κ^2 - Δ)^{α/2} u = \mathcal{W}, \;\; κ\in \mathbb{R}, \; α\in \mathbb{N}. \] of SPDEs. Using recent developements in the analysis of Discrete Exterior Calculus (DEC), we present a different, yet closely related, convergent GMRF approximation to these Matérn fields over complete, boundaryless Riemannian manifolds discretized as well-centered simplicial complexes. This convergent method (i) is agnostic to $α, κ$ and thus allows a universal approximation scheme for the precision and covariance matrices of the entire $(α, κ)$-family of GMRFs, so they may be inferred rather than guessed. (ii) inherently models pointwise and piecewise-smoothed measurements of a random field and approximates both equally well (iii) is computationally independent of the interpolants used - it suffers no overhead if one convergent interpolant were replaced with another suitable interpolant over the same mesh. Furthermore, we show that, on discretizations that are well-connected in a precise sense, and volume-concentrated, the precision matrices are spectral functions of a graph-laplacian. We provide a low rank approximator to the family of such Matérn GMRFs and mention a use case: reducing the number of measurements needed to model the GMRF by compressed-sensing.

Deep unfolding (DU) accelerates iterative optimizers by introducing learnable components and training them through unrolled iterations, but extending DU to the large-scale semidefinite programs (SDPs) common in robotics has remained limited. Unrolling a full-update conic solver such as COSMO exposes two obstacles that prior work on learned conic solvers has not: backpropagating through the per-iteration linear-system solve incurs memory quadratic in the problem size once the coefficient matrix is formed explicitly, and backpropagating through the positive semidefinite (PSD) cone projection becomes numerically unstable when eigenvalues coincide. We address the first obstacle with a matrix-free implicit differentiation rule that operates entirely through matrix-vector products, reducing memory from $O(n^2)$ to $O(n)$ and enabling backpropagation at scales where direct factorization runs out of memory. We address the second with a backward rule based on the Dalečkii--Krein representation of the Fréchet derivative, which remains well-defined under repeated eigenvalues. Together these make it possible to learn lightweight hyperparameter policies and warm-starts for a full-update conic solver. We evaluate on nonlinear covariance steering problems solved via sequential convex programming (SCP), as well as standalone SDPs and second-order cone programs ranging from max-cut and Lovász $\vartheta$ SDPs to robust estimation and control problems. The learned policies outperform state-of-the-art solvers across all problems, and can provide up to a 50$\times$ speedup depending on the class. When used as a subroutine in SCP, the learned approach delivers over a 30$\times$ speedup compared to COSMO.

We study training-free fixed-length descriptors for multivariate time series and ask not merely whether such a descriptor performs well, but when it can be expected to work at all. Our object of study is $D(τ)$, built from a time-lagged correlation matrix truncated at the Marchenko-Pastur edge so that only signal-bearing eigenvalues survive and classified by cosine similarity to class centroids with zero learned parameters. The central contribution is not the descriptor but a falsifiable applicability criterion for it. Working from a stationary Gaussian VAR(1) model, we argue that $D(τ)$ separates two classes when the signals are approximately stationary and the class information lives in their cross-channel temporal coupling rather than in marginal per-channel power. We derive, semi-formally, three consequences: a distinguishability condition, why the static ($τ=0$) covariance collapses to chance, and why a stationary but power-discriminated paradigm defeats the descriptor. The criterion is operational: a two-part pre-flight test -- an augmented Dickey-Fuller stationarity check and a power-baseline saturation check -- predicts applicability before any training. We validate both halves on a mixed assortment. On four paradigms that satisfy the criterion (Sleep-EDF, BCI-IV-2a, MIT-BIH, ESC-50) the descriptor is competitive with strong baselines at a fraction of their cost, reaching $88.5\pm4.5\%$ under 20-subject leave-one-subject-out on Sleep-EDF on a single CPU thread. On three that violate it -- non-stationary ERPs, and financial-volatility and wearable-stress regimes that are power-discriminated -- it fails exactly as the pre-flight predicts, and these negatives are the more informative half. We are explicit that $D(τ)$ is not the most accurate representation; its value is a compact, training-free embedding whose domain of validity is known in advance.

Estimating the probability that a treatment outperforms a control for an individual patient, called the Individual Probability of Treatment Benefit (IPTB), offers a clinically intuitive alternative to population-average metrics. However, existing methods for IPTB estimation are largely confined to binary treatment settings, despite the prevalence of dose-varying interventions in clinical practice. We propose a general framework for IPTB estimation with ordinal outcomes under discrete dose assignments, called Dose-AIPTB (Dose Attention-based IPTB). Our approach recasts the problem as binary classification over the unobserved sign of the individual treatment effect, constructing pseudo-labels from covariate-similar pairwise comparisons and aggregating them via attention mechanisms or Nadaraya-Watson kernel regression. This formulation naturally accommodates multiple discrete dose levels, extending beyond the binary treatment paradigm. Through numerical experiments on real-world and synthetic data under covariate shift, varying sample sizes, and heterogeneous outcomes, we demonstrate that attention-based aggregation consistently outperforms kernel alternatives. The framework provides a foundation for personalized dose selection grounded in individual-level benefit probabilities. Codes implementing the model are publicly available at https://github.com/NTAILab/AIPTBDose.

This thesis investigates how Bayesian principles can deepen our understanding of modern deep learning systems. While neural networks achieve remarkable predictive performance, their ability to generalize and to quantify uncertainty remains only partly understood. This thesis approaches this challenge from both methodological and theoretical angles: unifying Bayesian inference, function-space modeling, and large-deviation theory under a common probabilistic perspective. On the methodological side, the thesis introduces the Deep Variational Implicit Process (DVIP), a scalable Bayesian framework that extends implicit processes to deep architectures. Complementing this, two post-hoc methods -- the Variational Linearized Laplace Approximation (VaLLA) and the Fixed-Mean Gaussian Process (FMGP) -- are proposed to equip pretrained deterministic networks with calibrated uncertainty estimates. The theoretical contributions focus on one of the central open questions in modern machine learning: why do large, over-parameterized neural networks generalize so well? To address this, the thesis develops a unified probabilistic framework that connects three key mechanisms -- diversity, smoothness, and stochasticity -- within the language of PAC-Bayesian and large-deviation theory.

World models in robot learning predict future states from visual observations and actions, enabling agents to reason about the consequences of their controls. However, many action-conditioned models are evaluated in settings where motion is dominated by immediate control, whereas aquatic surface vehicles and other real-world objects continue moving under inertia and are displaced by hidden ambient drift, such as water currents or wind. We propose FlowMo-WM, an end-to-end trainable visual world model that infers object-centric motion state and a predictive long-history context associated with hidden drift from image-action histories without direct supervision of flow fields. FlowMo-WM factorizes image-action history into a short-history latent state, trained to summarize object-centric motion, and a longer-history context, trained to summarize slowly varying exogenous influences. A zero-context residual transition separates action-conditioned base dynamics from context-dependent drift effects during latent rollout. In simulated aquatic surface-vehicle environments with diverse hidden flows, disturbances, and randomized vehicle dynamics, FlowMo-WM improves long-horizon rollout accuracy over representative action-conditioned latent world models. Prediction-time context ablations, in which the inferred context is zeroed or shuffled during rollout, show that the ambient context is important for stable prediction under hidden drift, while frozen linear probes characterize information encoded in the learned factors.

Strategic reasoning under uncertainty underpins consequential decisions in negotiation, finance, and policy, but prevailing game-play benchmarks collapse heterogeneous reasoning dimensions into a single scalar, leaving the capability structure of frontier LLMs unexamined. We introduce Poker Arena, a no-limit Texas Hold'em tournament platform that couples a three-layer memory architecture (within-hand, session, and cross-session) with a nine-axis cognitive profile decomposing strategic reasoning into interpretable dimensions such as bet-sizing calibration and positional awareness. We evaluate seven frontier models across 50 sessions of 1,000 hands and a controlled memory ablation; tournament chips and aggregate axis score order the field differently: Claude Opus 4.6 wins +$15,730 chips with 14 first-place finishes, yet ranks only fifth of seven on mean axis score, while persistent memory helps some models and hurts others. These findings show that multi-axis evaluation surfaces capability structure that scalar leaderboards systematically misrank, with cross-dimensional consistency outweighing peak performance on any single axis.

Can Large Language Models (LLMs) understand and reason about quantum operators? Despite their remarkable capabilities in mathematics and symbolic reasoning, LLMs remain inherently blind to quantum representations such as unitary matrices. In this work, we take a step toward bridging this gap by introducing an approach that maps unitary operators into the latent space of an LLM, enabling unified modeling over quantum and linguistic inputs. We instantiate this idea on Clifford+T circuit synthesis over a Pauli rotation gate set, where our model achieves results competitive with state-of-the-art methods and scales consistently with training data, with no signs of saturation. Our approach further enables language-conditioned synthesis, allowing gate constraints unseen during training to be specified directly in natural language. This work suggests a path toward quantum--aware foundation models that can natively interpret and reason about quantum operations, which could have broader implications reaching across quantum compilation and algorithm discovery.

Current cultural alignment approaches focus on inference-time interventions, assuming models already contain sufficient cultural knowledge. We argue modern LLM pipelines suffer from a cultural data funnel. Using a multidimensional tagging framework across pretraining, fine-tuning, alignment, and reasoning datasets, we show explicit cultural signals decline sharply during post-training, while geographically concentrated, task-specialized data dominates. Multilinguality enhances geographic diversity of cultural knowledge but does not ensure balanced representation. Our tags improve downstream cultural benchmark performance, demonstrating that advances require shifting focus in training data pipelines. To facilitate future research, we release our culturally tagged dataset with 5.6M samples at https://huggingface.co/datasets/CohereLabs/CultureMarkers.

Manual testing, in which testers follow natural language instructions to validate system behavior, remains essential for uncovering issues that are difficult to capture with automation. However, manual test cases often contain test smells, quality issues such as ambiguity, redundancy, or missing checks that reduce reliability, maintainability, and reproducibility. Existing detection approaches largely depend on manually engineered rules and thus struggle to generalize and scale across heterogeneous test suites. In our previous work, we assessed the feasibility of using Small Language Models (SLMs) for test smell detection by evaluating GEMMA-3-4B, LLAMA-3.2-3B, and PHI-4-14B on test steps from 143 real-world Ubuntu test cases, covering seven smell types. PHI-4-14B achieved the best performance. In this article, we investigate whether a contemporary Large Language Model (GEMINI-3-PRO-PREVIEW) available at the time of the study can identify test smells in natural language manual test cases using a prompt-based, whole-test-case analysis strategy. Unlike approaches that analyze individual test steps in isolation, our approach evaluates complete test cases, enabling the model to consider relationships and dependencies among test steps. We evaluate the approach on 100 Ubuntu test cases covering seven test smell types and compare its performance against previously evaluated SLMs, including GEMMA-3-4B, LLAMA-3.2-3B, and PHI-4-14B. Our results show that GEMINI-3-PRO-PREVIEW outperforms the SLMs, while producing actionable explanations that can help practitioners revise manual test cases for greater clarity and consistency. We also find that test smells are pervasive in practice, with nearly one detected test smell per step on average, highlighting the need for scalable and automated quality support for manual testing artifacts.

Enforcing functional inequality constraints such as monotonicity and convexity in neural networks is a fundamental challenge in many industrial and scientific applications. Classical one-sided penalty methods, along with primal-dual methods gated by complementary slackness, provide constraint gradients only at violated locations, resulting in fragile satisfaction. Architectures that guarantee feasibility by construction, on the other hand, remain largely limited to elementary cases and impose additional inductive biases. We introduce neural slack variables, a deep learning native primal-side approach that converts constraint enforcement into a regression problem by coupling the primary network with a jointly learned auxiliary network. The auxiliary network serves as a valid target for the primary network's constraint quantities, inducing feasibility and regularity. Neural slack variables achieve zero measured violations on dense-grid monotonicity and convexity test cases, where penalty and primal-dual baselines leave residual violations, and enable arbitrage-free learning of volatility surfaces, an open industrial challenge in quantitative finance.

Predictive code completion greatly accelerates how quickly developers work. In spreadsheets, despite being much more common, such auto-completion features are virtually non-existent. To address this gap, we introduce a benchmark for systems that observe a sequence of user actions in a spreadsheet and predict future actions. Two challenges are (1) the absence of edit histories in public spreadsheet corpora and (2) the complex space of spreadsheet actions (spatial, temporal, composite). To address (1), we manually curate 52 sequences of 12K actions that recreate spreadsheets from public corpora, seeded by parametrized heuristics and LLM refinement. To address (2), we propose an online evaluation that expects a prediction after each user action, accepts or rejects that prediction, updates the future actions upon acceptance, and repeats this until the target spreadsheet is obtained. We use multiple baseline predictors (including zero-shot LLMs, fine-tuned SLMs, and classical models) and analyze different properties that our benchmark teaches us, including but not limited to: properties of saved actions and false positives, efficiency, effect of user profiles, effect of triggers, and effect of context.

Neural responses in cortex exhibit substantial trial-to-trial variability in response to repeated stimuli, while peripheral sensory neurons respond far more consistently, leading many to wonder whether stochasticity may carry meaning. Existing work has argued that noise and signal correlations may be optimized for discrimination in animals, whereas artificial neural network (ANN) studies have shown similar benefits of noise in machine learning tasks, although most ANN work has neglected the effects of correlations. Here we investigate whether correlated noise improves the robustness of artificial neural networks to adversarial attacks and naturalistic image modifications. Using the covariance of activations under modified versus clean inputs, we find that structured noise may significantly improve network robustness. Robustness to naturalistic image modifications benefits most from structure, but this structure transfers poorly across modification types. In contrast, noise structure from adversarial attacks can generalize to other kinds of attacks. These results suggest that structured noise in ANN activations generally improves robustness, establishing a biologically plausible strategy for creating robust artificial neural networks that only relies on local information.

Finding the shortest program that generates a sequence is uncomputable, and for six decades that fact has been mistaken for a wall around finding any generating program. It is not a wall but a price, and this paper measures it. For every algorithm that learns about a candidate program only through its score, a class spanning Levin search, evolutionary methods, simulated annealing, and the cross-entropy method, we define the coupling width of a search problem and prove an unconditional worst-case lower bound, exponential in that width with base one less than the domain size. From it follows a conservation law: structural knowledge injected into a search trades one for one against the search it removes, and their sum can never fall below the length of the program sought. Levin's 1973 upper bound and the lower bound proved here are the two ends of one conserved quantity, closing on each other as the instruction set grows. The only escape is to read a candidate's structure rather than its score, and its price, which we prove for generic targets, is incompleteness. A deterministic engine built on this theory recovers a generating program, certified by compressing its data and predicting an unseen continuation, for 2,383 of 3,914 sequences across four independent populations, including 244 of the 256 elementary cellular automata, with measured discovery cost rising along program length more than an order of magnitude inside the score-oracle worst case.

Recursive training of generative models on their own outputs can lead to model collapse, a compounding drift away from the true data distribution. Existing theoretical works bound finite-round error accumulation in the context of diffusion models, but two questions remain open:~what distribution does the recursion converge to, and how fast? We answer both, isolating a mechanism distinct from imperfect learning: even with perfect score estimation and exact sampling, the early stopping of the reverse diffusion (required for numerical stability) drives a progressive drift away from the data distribution. We prove that this recursion converges geometrically to a unique limiting distribution, which admits a closed-form characterization as an infinite mixture of increasingly Gaussian-smoothed versions of the data distribution. A Hermite spectral decomposition of this limit reveals that recursive training acts as a low-pass filter: higher-order modes, which encode fine non-Gaussian structure, are attenuated much more strongly than coarse modes. This spectral picture motivates annealed truncation schedules that progressively shrink truncation times across retraining rounds; we prove that any schedule converging to $0$ asymptotically eliminates recursive compounding. Finally, we show our idealized characterization is robust: in the presence of discretization and score estimation errors, the learned distribution remains in a Wasserstein-2 ball around the ideal limit, with mode-dependent contraction rates that contract high-order errors faster than low-order ones. We validate the theory on synthetic Gaussian mixtures and CIFAR-10.

RL post-training has become increasingly pivotal for improving diffusion policies, but existing diffusion policy-gradient methods are often unstable and cannot achieve reliable policy improvement. We identify the cause as the double-drift phenomenon: optimizing a variational surrogate can let the ELBO separate from the true log-likelihood, which then makes the resulting proxy policy gradient misaligned with the true policy gradient of expected return. We propose \textbf{DiPOD}, a diffusion policy optimization framework that maintains tight-bound behavior throughout training by interleaving self-distillation with policy-improving gradient updates. This leads to a simple and practical algorithm: augmenting each diffusion policy-gradient update with an on-policy ELBO regularizer. Across diffusion language model post-training and continuous-control diffusion policies, DiPOD substantially stabilizes training and reaches higher rewards than previous methods.

Nonlinear dynamics and the strong couplings that arise between multiple effectors undermine the assumptions behind conventional, linear control allocation techniques. When flight enters regimes where nonlinear effects dominate, linear allocators exhibit reduced accuracy due to increased model mismatch, which subsequently degrades performance and robustness of the flight control system. High fidelity onboard models and black box data driven approaches can recover accuracy across the flight envelope, but respectively impose computational burdens prohibitive for real time allocation and sacrifice the interpretability required for verification and fault diagnosis. This paper addresses these limitations by learning an explicit, physics constrained analytical model of the control effectiveness mapping from representative flight data using Sparse Identification of Nonlinear Dynamics. The resulting mapping is compact, interpretable, and admits analytical derivatives, enabling efficient computation within nonlinear solvers that additionally incorporate actuator dynamics, without requiring an onboard model. An online adaptation mechanism monitors prediction residuals and refreshes the model when significant plant changes are detected, providing graceful reconfiguration under actuator failures and varying operating conditions. The methodology is evaluated on a high fidelity nonlinear benchmark aircraft across a range of aggressive maneuvers, achieving accuracy comparable to a full nonlinear onboard model while substantially reducing computational cost relative to established baselines.

Large Language Models (LLMs) have made notable progress in automated theorem proving, yet existing formal benchmarks remain limited in both mathematical coverage and difficulty. Most are concentrated in areas that are easier to formalize, such as algebra and elementary number theory, and provide limited coverage of subfields that require deeper reasoning, including mathematical analysis. To address this gap, we introduce MA-ProofBench, to the best of our knowledge, the first formal theorem-proving benchmark dedicated to Mathematical Analysis. The benchmark contains 200 formalized theorems covering 6 core topics and 27 subcategories, including measure and integration theory, complex analysis, and functional analysis. The problems are divided into two difficulty levels, an undergraduate level (Level I, 100 problems) and a Ph.D. qualifying level (Level II, 100 problems), to evaluate how well LLMs perform formal reasoning at different mathematical depths. Each problem is constructed through a human-led, LLM-assisted formalization pipeline followed by independent expert review, ensuring that the formal statements remain faithful to the original mathematics. We evaluate a range of recent general-purpose reasoning models and formal theorem provers on MA-ProofBench. However, most models perform poorly: even the best-performing model, GPT-5.5, achieves only 16% Pass@8 on Level I and 5% on Level II, while most models stay close to 0% on Level II. Further analysis identifies Mathlib hallucinations and incomplete proofs as the two dominant failure modes, while an evaluation on the natural-language version of the benchmark exposes a clear gap between informal and formal reasoning. MA-ProofBench is intended to serve as a reliable reference for tracking progress in formal mathematical reasoning in advanced domains.

Machine-learned anomaly detection is reshaping searches for new physics, but it has outrun the statistics used to interpret it. A raw anomaly score has no calibrated meaning, a model that scans many regions inflates the look-elsewhere effect, and the asymptotic significances the field relies on are blind to the background mismodelling that anomaly detectors are especially prone to. We propose a calibration layer, built on conformal prediction, that turns any anomaly score into a defensible significance with distribution-free, finite-sample guarantees. Conformal prediction converts scores into valid local p-values, weighted and Mondrian variants repair the sideband-to-signal-region exchangeability failures that resonant searches suffer, and a Gross-Vitells step carries the result through to a look-elsewhere-aware global significance. The layer does two things at once. It exposes miscalibration that the standard pipeline cannot see, and it corrects it without retraining the detector. On public LHC Olympics data, a classifier develops a substructure-mass correlation that makes sideband-calibrated background p-values anti-conservative. Taken at face value, this manufactures a $\sim 46σ$ excess from background sculpting alone, which the label-free weighted correction removes, restoring an honest null. When run as a blind wide-mass bump hunt, the standard asymptotic and unweighted procedures fabricate $\gtrsim10σ$ excesses and $\approx5σ$ excesses even in signal-free windows, while the conformal layer raises no false alarms and its global false-positive rate is verified on background-only pseudoexperiments. The result is an auditable, detector-agnostic path from an uncalibrated score to a trials-factor-aware significance, ready to be folded into experimental anomaly searches.

Large language model (LLM) agents have achieved strong performance on a wide range of benchmarks, yet most evaluations assume static environments. In contrast, real-world deployment is inherently dynamic, requiring agents to continually align their knowledge, skills, and behavior with changing environments and updated task conditions. To address this gap, we introduce EvoArena, a benchmark suite that models environment changes as sequences of progressive updates across terminal, software, and social domains. We further propose EvoMem, a patch-based memory paradigm that records memory evolution as structured update histories, enabling agents to reason about environmental evolution through changes in their memory. Experiments show that current agents struggle on EvoArena, achieving an average accuracy of 39.6% across evolving terminal, software, and social-preference domains. EvoMem consistently improves performance, yielding an average gain of 1.5% on EvoArena and also improving standard benchmarks such as GAIA and LoCoMo by 6.1% and 4.8%. Beyond individual tasks, EvoMem further improves chain-level accuracy by 3.7% on EvoArena, where success requires completing a consecutive sequence of related evolutionary subtasks. Mechanistic analysis shows that EvoMem improves evidence capture in the memory, indicating better preservation of complete evolving environment states. Our results highlight the importance of modeling evolution in both evaluation and memory for reliable agent deployment.

World models that capture how actions induce physical change enable scalable robot learning without reliance on embodiment-specific action labels. Pixel-space video models provide broad visual priors but expend model capacity on dense appearance reconstruction, while direct action models require embodiment-specific labels that hinder scalability. We present $μ_0$, a scalable world model based on 3D traces. Rather than predicting dense pixels or directly modeling actions, $μ_0$ forecasts smooth 3D trajectories for salient interaction points such as objects, tools, hands, and contact regions, yielding a compact, embodiment-agnostic motion interface. To enable training from diverse video sources, our TraceExtract system automatically extracts 3D supervision by selecting keypoints, constructing globally aligned traces, and associating motion segments with hierarchical language captions. This TraceExtract supervision pretrains $μ_0$ by combining a pretrained vision-language backbone with a modular trace expert, which represents each query via B-spline control points and predicts future traces. Experiments show that $μ_0$ outperforms baselines in both 2D and 3D trace prediction, including trace prediction models and tokenized VLM methods. Because $μ_0$ is frozen and reusable, it can be paired with action experts for downstream robot embodiments. Despite action-free pretraining, the resulting trace-conditioned policies achieve performance competitive with VLA models pretrained with action supervision, such as $π_0$. These results establish 3D traces as a scalable and transferable representation for cross-embodiment manipulation.

Retrieval-augmented generation (RAG) has become a standard mechanism for grounding language models in external knowledge, yet conventional retrieval based on lexical or semantic similarity is poorly suited for complex reasoning tasks: a semantically similar problem may demand an entirely different solution strategy, while a superficially different problem may share the same underlying reasoning pattern. We propose Retrieval-Augmented Reinforcement Fine-Tuning (RA-RFT), a post-training framework that teaches language models to reason by analogy. RA-RFT uses gold-relevance distillation to train a retriever that ranks contexts by expected reasoning benefit rather than semantic overlap, and then fine-tunes the policy model via reinforcement fine-tuning methods with retrieved analogous demonstrations, so the model learns to leverage reasoning traces under verifiable outcome rewards. We further analyze the diversity of retrieved contexts and find that reasoning-aware retrieval surfaces complementary solution strategies that provide distinct reasoning scaffolds for individual problems. Across challenging mathematical reasoning benchmarks, RA-RFT consistently outperforms standard reinforcement fine-tuning methods. For example, it improves AIME 2025 average@32 accuracy by 7.1 and 2.8 points over GRPO for Qwen3-1.7B and Qwen3-4B respectively -- suggesting that reasoning-aware retrieval is a complementary axis of improvement and orthogonal to advances in reward design or training curricula.

Articulated tool manipulation remains a major challenge in dexterous robotics due to the need to coordinate internal degrees of freedom and contact-rich interactions. While prior work has largely focused on rigid objects, articulated tool use remains underexplored because of its physical complexity and the difficulty of learning functional grasping and manipulation policies. We present Mana (Manipulation Animator), a general sim-to-real framework that reinterprets dexterous manipulation as an animation problem. Inspired by computer animation, Mana employs a coarse-to-fine pipeline that transforms procedurally-generated grasp keyframes into manipulation trajectories through motion planning and reinforcement learning. The data generation process is largely automatic, requiring only a few mouse clicks to specify functional affordances (<1 minute per tool). Across four articulated tools spanning different scales and joint types, Mana achieves zero-shot sim-to-real transfer for both grasping and in-hand manipulation, demonstrating a scalable approach to dexterous articulated tool use.

Spatial reasoning, the ability to determine where objects are, how they relate, and how they move in 3D, remains a fundamental challenge for vision-language models (VLMs). Tool-augmented agents attempt to address this by augmenting VLMs with specialist perception modules, yet their effectiveness is bounded by the action interface through which those tools are invoked. In this work, we study how the design of this interface shapes the agent's capacity for open-ended spatial reasoning. Existing spatial agents either employ single-pass code execution, which commits to a full analysis strategy before any intermediate result is observed, or rely on a structured tool-call interface that often offers less flexibility for freely composing operations or tailoring the analysis to each task. Both designs offer limited flexibility for open-ended, complex 3D/4D spatial reasoning. We therefore propose SpatialClaw, a training-free framework for spatial reasoning that adopts code as the action interface. SpatialClaw maintains a stateful Python kernel pre-loaded with input frames and a suite of perception and geometry primitives, letting a VLM-backed agent write one executable cell per step conditioned on all prior outputs, enabling the agent to flexibly compose and manipulate perception results and adapt its analysis to both intermediate text and visual observations and the demands of each problem. Evaluated across 20 spatial reasoning benchmarks spanning a broad range of static and dynamic 3D/4D spatial reasoning tasks, SpatialClaw achieves 59.9% average accuracy, outperforming the recent spatial agent by +11.2 points, with consistent gains across six VLM backbones from two model families without any benchmark- or model-specific adaptation.

Cinematic video depicts multiple subjects acting or interacting at specific moments, captured with deliberate camera movement, and stitched together by shot transitions. Together, these elements demand a level of fine-grained control beyond current text-to-video models. Existing work addresses each axis in isolation: multi-subject personalization, temporal control, multi-shot synthesis, or camera control; no prior framework jointly integrates all four. We present CineOrchestra, a unified video diffusion model that controls subjects, events, cameras, and shot transitions simultaneously. Our key insight is that these heterogeneous cinematic elements share a fundamental structure: each is an entity acting over a specific temporal interval, which can therefore all be expressed through one shared structure of entity-centric conditioning primitives, augmented with reference images for visual entities. This formulation reduces the architectural challenge to a single positional encoding problem, which we solve with two parameter-free coordinated rotary embeddings: (a) an interval-sampled temporal RoPE that yields consistent attention behavior across events of dramatically varying duration, and (b) a 2D entity-temporal cross-attention RoPE that disambiguates per-entity conditions and routes each to its corresponding spatiotemporal region. On two new benchmarks, CineOrchestra outperforms six per-axis specialists on dense caption following and shot-transition timing, with consistent gains in a pairwise user study and component ablations.

Low-rank adaptation (LoRA) and its variants provide a memory- and compute-efficient alternative to full fine-tuning of pre-trained models. However, questions remain about the comparative generalizability of these approaches and how the structural restrictions on low-rank updates preserve effective adaptation performance. We present a historical framing, covering the past (full fine-tuning and original LoRA), the present (different variants of LoRA), and propose simpler, cheaper, parameter-efficient extensions by inducing sparsity within existing LoRA variants: Cheap LoRA (cLA), training a single low-rank factor with the other fixed (deterministically or, in its randomized variant, stochastically), and the chained circulant variant, ${c}^3$LA. We frame cLA as a structured instance of asymmetric LoRA, serving as a controlled column-subspace restriction of full fine-tuning. We derive information-theoretic generalization error bounds for these variants, marking one of the first endeavors in this area. Empirically, we evaluate 11 fine-tuning methods across 10 pre-trained models and 14 datasets, analyzing the fine-tuned models' performance and generalization using tools such as loss landscapes and spectral analysis. Despite the sensitivity of fine-tuned models to the pre-trained model, datasets, and other factors, our study suggests that restricting LoRA-based PEFT methods' adaptation to a sparse, structured column space remains competitive across tasks with their parameter-matched baselines while reducing up to 10% training time and peak GPU memory up to 15%, even with a naïve, non-optimized, sparse implementation. Our theoretical and empirical generalization measures provide a more consistent and principled approach to their cost-effective adaptation than commonly used analytical tools. Overview and code are available at: https://elicaden.github.io/Beyond_LoRA/.

Positional encodings (PEs) enhance the power of graph neural networks (GNNs), both theoretically and empirically. Two of the most popular families of PEs - spectral (e.g., Laplacian eigenspaces, effective resistance) and walk-based (polynomials of the adjacency matrix) - are theoretically equivalent in expressive power, with expressivity between the 1-WL and 3-WL tests. However, this equivalence assumes the GNN uses the "complete" version of these PEs, which requires $O(n^3)$ time and space complexity. Instead, practitioners commonly use truncated variants of these encodings, such as the first $k$ eigenspaces or powers of the adjacency matrix. However, the theoretical properties of these truncated PEs are unknown. In this work, we initiate the study of these truncated PEs. Theoretically, we show that, under truncation, several families of PEs are fundamentally different in expressive power. As a corollary, we show that truncated spectral PEs are no longer stronger than the 1-WL test. We also study a family of spectral PEs, the $k$-harmonic distances, to highlight the differences in expressive power of even closely related truncated PEs. Finally, we experimentally show that a mix of truncated PEs is preferable to any single family on real-world datasets.

Reproducibility in the social and behavioral sciences is typically evaluated by independent researchers who reanalyze the original data to assess whether the published findings can be recovered. However, such approaches are resource-intensive and difficult to scale. Here, we show that large language models (LLMs) can automate reproducibility assessments. Using N=76 published studies with predefined claims from the behavioral and social sciences, we compare LLM-generated analysis with the original findings and human reanalysis. For 7 studies, the LLM could not produce a viable effect size estimate. For the remaining studies, our LLM pipeline recovered the original effect sizes in 41% of studies using a +/-0.05 tolerance in Cohen's d. Further, our LLM pipeline reached the same qualitative conclusion as the original study in 96% of cases, where conclusions indicate whether the reanalysis supports the original claim. For comparison, human reanalysts recovered the original effect sizes in 34% of studies and reached the same qualitative conclusion in 74% of cases. Together, these results show that LLMs can serve as a scalable tool for automated reproducibility assessment and provide a foundation for systematic auditing of empirical results in the social and behavioral sciences.

Current LLM-based research agents have advanced through agent orchestration, yet largely overlook scientific knowledge orchestration. Existing works often reduce papers to abstracts, surface mentions, and flat \texttt{cites} edges, omitting key entities, claims, evidence, mechanisms, and method lineages essential for scientific reasoning. To this end, we introduce \textbf{Agents-K1}, an end-to-end knowledge orchestration pipeline that converts raw documents into agent-native scientific knowledge graphs. Agents-K1 integrates three components under a unifying theoretical foundation: a multimodal parser whose five-module schema captures entities, multimodal evidence, citations, and typed inter-entity relations across the full paper rather than abstracts alone; a 4B information-extraction backbone trained with GRPO under a rule-based reward; and a graphanything CLI, a tri-source agent interface that unifies web search, multimodal graph retrieval, and cross-document traversal. On top of this, we process 2.46 million scientific papers across six subjects to produce \textbf{Scholar-KG}, of which we release a one-million-paper subset, and the full Scholar-KG is accessible via the SCP link below. The same pipeline can be extended to general-domain corpora and to schema-conformant data synthesis. Extensive experiments demonstrate that Agents-K1 achieves superior performance in scientific information extraction, knowledge graph construction, and multi-hop scientific reasoning.

With the growth of LLMs' (Large Language Models) capabilities, there has been an increasing push to curate high quality datasets by filtering samples in the training data. In general, Data Attribution (DA) methods aim to estimate how individual samples in a training dataset can precondition a model to generate certain outputs. As an example, one might be interested in which samples in the data could be the source of toxic behavior after training the LLM. Many methods quantify this conditioning through the paradigm of influence functions. While methods of this family are effective in its function, they lack the necessary processing speed and storage compactness to be practically implemented on large datasets. We propose a method, Influcoder, as a quick and cost-effective approach to influence-based Data Attribution at scale.

Tool-augmented LLM agents commonly rely on step-wise atomic tool calls, where each invocation, observation, and value transfer is exposed in the main reasoning trace. This creates an \emph{execution-granularity mismatch}: locally deterministic tool workflows are unfolded into repeated model-visible decisions, consuming context and forcing the model to manage low-level dataflow in the trace. We introduce \textbf{HyperTool}, a unified executable MCP-style tool interface that changes the model-visible unit of tool execution. A model invokes HyperTool with a code block that can call existing tools through their original schemas, manipulate returned values, and pass intermediate results locally, folding deterministic tool subroutines into a single outer call. To train models to use this interface, we synthesize HyperTool-format trajectories from cross-tool compositional tasks and verify them in real MCP environments. On MCP-Universe, HyperTool improves average accuracy from 15.69\% to 35.29\% on Qwen3-32B and from 9.93\% to 33.33\% on Qwen3-8B, and surpass GPT-OSS and Kimi-k2.5 on average accuracy, showing that our HyperTool can substantially improve multi-step tool use.

LLM-based agents have shown increasing potential in automating scientific discovery. Given an optimizable metric and an execution environment, they can propose, validate, and iterate scientific solutions, and have produced results that outperform human-designed approaches. As model capabilities continue to improve, we argue that the bottleneck for autonomous scientific discovery is shifting from prescribing agent workflows to designing agent environments: the resources, constraints, and interfaces that shape agent behavior. We frame this as environment engineering: building environments that amplify productive behaviors, such as open-ended exploration, systematic artifact management, and inter-agent collaboration, while suppressing harmful behaviors, such as reward hacking and high-friction human oversight. We present EurekAgent, an environment-engineered agent system for metric-driven autonomous scientific discovery. EurekAgent engineers the environment along four dimensions: permissions engineering for bounded agent execution and isolated evaluation; artifact engineering for filesystem and Git-based collaboration; budget engineering for budget-aware exploration; and human-in-the-loop engineering for easy human supervision and intervention. EurekAgent sets new state-of-the-art results on multiple mathematics, kernel engineering, and machine learning tasks, including new state-of-the-art 26-circle packing results discovered with less than $11 in total API cost. We open-source our code and results, and call for environment engineering as a core research direction for developing reliable autonomous research agents.

To test and debug hardware modules, it is common to write two programs: a driver, which translates high-level transactions into interactions on the module's input and output signals, and a monitor, which analyzes a signal-level execution trace and recognizes a transaction. These two programs are commonly implemented separately for each hardware protocol, but this separation entails manual effort and risks inconsistencies. We advocate an alternative approach. We present a DSL in which users specify hardware communication protocols as succinct imperative programs. Crucially, the same specification can be used to both drive designs and monitor transactions. We present the design of a tool, which given a specification in our DSL and a waveform, automatically infers a transaction-level trace consistent with the waveform. We discuss plans to evaluate our DSL on real-world interconnects such as Wishbone and AXI-Stream.

This paper examines three recent frameworks for understanding the cognitive and epistemic consequences of artificial intelligence: Tri-System Theory, Thinkframes, and System 0. It argues that while the first two capture important dimensions of AI's influence on individual reasoning and collective epistemic practices, System 0 occupies a theoretically distinctive position that neither can fully replicate. The paper introduces the concept of cognitive colonization, according to which AI systems can embed external interests within the architecture of the self in ways that are difficult for users to perceive. Because such systems are already widely deployed, understanding these invisible forms of influence is an urgent philosophical and practical task.

On-policy distillation (\textsc{OPD}) has recently become a prominent post-training recipe by combining two desirable ingredients: on-policy student trajectories and dense teacher supervision. However, how this hybrid changes a model's parameters remains unclear. Across several language and vision-language model pairs and \textsc{OPD} use cases, our analysis yields two main findings. On sparsity, \textsc{OPD} updates are small and coordinate-sparse. They are distributed across layers, with the largest relative movement usually appearing in FFN modules. This sparse structure is operationally useful: training only the discovered subnetwork nearly recovers full-training performance. The sparse support does not remove the need for adaptive optimization: SGD, previously reported to be competitive in \textsc{RLVR}, underperforms AdamW in our \textsc{OPD} optimizer ablation, suggesting that dense teacher supervision preserves useful momentum structure and heterogeneous second-moment scales. On geometry, the updates are numerically full-rank but spectrally concentrated; they lie mostly away from the principal singular subspaces of the source weights and fall disproportionately on coordinates where the source weights are close to zero. These findings suggest that dense teacher supervision does not turn \textsc{OPD} into ordinary dense parameter rewriting; instead, \textsc{OPD} retains important geometric signatures of on-policy post-training.

Detecting LLM reasoning failures at inference time without ground-truth labels has motivated a wide range of confidence baselines, including self-consistency, semantic entropy, and P(True), built on within-question sampling and self-evaluation. Operad theory, the formalism for systems built by iterated substitution, suggests a complementary diagnostic: a model's direct answer to a compositional query should agree with the answer it produces by composing a stated decomposition of the same query. We instantiate this idea as operadic consistency (OC), a per-question signal. Across twelve instruction-tuned LLMs (4B to 671B parameters, open-weights and closed-source) on four multi-hop QA datasets, OC is strongly correlated with accuracy on every dataset (Pearson $r \in [0.86, 0.94]$, all $p \leq 0.0004$), and is the only signal we evaluate with $r \geq 0.85$ uniformly across all four datasets. Chain-of-thought self-consistency (CoT-SC; Wang et al., 2023) matches OC on HotpotQA and DROP ($r = 0.93, 0.87$) but drops to $r \approx 0.45$ on MuSiQue and StrategyQA. At the per-question level, OC contributes information beyond CoT-SC and semantic entropy on every dataset (cluster-robust $p \leq 10^{-16}$ for the OC coefficient), and the conclusion is robust to additionally controlling for constructed decomposition-aware baselines ($p \leq 10^{-13}$). The same signal yields selective-prediction improvements (accuracy at fixed coverage) over a tuned CoT-SC baseline at the equal-cost $K = 3$ budget (AUARC lifts of +0.086 to +0.096 and AUROC lifts of +0.092 to +0.164; 95% CIs exclude zero on every cell). On five frontier thinking models, where the decomposition is extracted from the model's own chain of thought, the same equal-cost comparison gives positive selective-prediction point-estimate lift on all 16 (dataset, budget, metric) cells tested, with 95% CIs excluding zero on 12 of the 16.

We introduce SkMTEB, the first comprehensive MTEB-style text embedding benchmark for Slovak, a low-resource West Slavic language, comprising 31 datasets across 7 task types -- nearly 4$\times$ the depth of existing multilingual benchmark coverage for Slovak. Our evaluation of 31 embedding models reveals that large instruction-tuned multilingual models achieve the strongest performance, while existing Slovak-specific models trained for NLU tasks transfer poorly to embedding tasks. To address the need for efficient, locally-deployable Slovak embeddings, we develop \texttt{e5-sk-small} (45M parameters) and \texttt{e5-sk-large} (365M) by applying vocabulary trimming and fine-tuning to Multilingual E5 models. Despite size reductions of up to 62\%, our open-source models achieve competitive performance with proprietary APIs while remaining locally deployable for semantic search and retrieval-augmented generation (RAG). We release the benchmark, models, datasets, and code openly, hoping our approach offers a replicable path for other under-resourced languages.

Recursive language models (RLMs) showed that recursion over model calls is an effective strategy for long-context reasoning, and production coding agents have begun to write code that spawns subagents at scale, most recently in Anthropic's dynamic workflows. We name and study the pattern between these two lines of work, where the recursive unit is a full agent harness with filesystem tools, code execution, and planning rather than a model call with no tools. We call this the Recursive Agent Harness (RAH) and frame it as harness recursion, the code-first extension to the model recursion of RLMs. A parent agent generates and runs an executable script that spawns subagent harnesses in parallel for fine-grained workloads and uses structured function calls for small subtasks. We provide a controlled evaluation on long-context reasoning. With the backbone held fixed at GPT-5 to match the published Codex and RLM baselines, RAH improves the Codex coding-agent baseline from 71.75% to 81.36% on Oolong-Synthetic (199 samples, 13 context-length buckets up to 4M tokens), a gain attributable to the harness rather than the model. With a stronger backbone, Claude Sonnet 4.5, the same design reaches 89.77%.

Recent growth in compute power has made it increasingly feasible to use large-scale agent-based models to simulate complex adaptive systems. A central difficulty is that such models contain many local rules and parameters, where small changes can lead to runaway behaviour, population collapse, or saturation at artificial bounds. We study this problem in a continuous predator-prey system where sheep and wolves are active agents with local sensing, internal energy, and recurrent neural network-based controllers. We ask whether environmental and demographic parameters can be tuned so that the resulting population dynamics resemble classical Lotka-Volterra cycles. We optimise these parameters with a feature-based loss that rewards sustained oscillations, phase lag, bounded populations, and long-term persistence, first for random controllers and then for evolved controllers in a more naturalistic setting. The model is implemented in ABMax, a JAX-based agent-based modelling framework that enables efficient batched simulation on hardware accelerators.

Catastrophic forgetting is often viewed as the destruction of previously learned knowledge during sequential learning. Building on the Accessibility Collapse framework, we investigate the geometric structure of recoverability in continual learning. Using Split CIFAR-100 and a sequentially trained ResNet-18, we analyze recoverability, representational drift, and recovery complexity across ten tasks. We introduce Recovery Subspace Dimensionality (k_t), a measure of the minimum number of singular directions required to preserve 90 percent of full probe performance. Contrary to our Recoverability Diffusion hypothesis, recovery dimensionality remains stable throughout training (mean k_t = 8.0) despite substantial representational drift. Principal-angle drift strongly predicts recoverability (r = -0.862), and a simple geometric model explains 82.2 percent of recoverability variance. These findings support the Stable Recovery Manifold hypothesis, suggesting that forgotten knowledge remains compactly decodable despite representational reorganization. The results indicate that catastrophic forgetting is primarily an accessibility and manifold-alignment problem rather than information destruction.

Question decomposition, i.e. breaking a complex query into simpler sub-queries whose answers are composed to produce a final answer, is a widely used strategy for improving LLM reasoning, yet it currently lacks a rigorous mathematical foundation. In this paper, we propose operads, mathematical structures that model many-in, one-out operations and compositions thereof, as a natural framework for describing question decomposition. We define the questions operad $Q$, in which operations correspond to question templates and composition corresponds to substitution of sub-answers, and show how QA models can be interpreted as algebras over $Q$. Beyond reframing existing practice, this operadic perspective points toward new methods, in particular a notion of operadic consistency, which measures whether a QA model's answers agree across the partial collapses of a question decomposition tree. Empirical evaluation of operadic consistency is reported in our companion paper (Bottman, Liu, and Richardson, 2026), which finds it strongly correlated with accuracy across twelve LLMs and four multi-hop QA datasets and outperforming standard temperature-based self-consistency baselines. We argue that operads are the natural mathematical home for question decomposition, and that invariants such as operadic consistency open new directions for analyzing and improving the reliability of multi-step reasoning.

Aerial wildfire suppression requires not only predicting fire spread, but also designing effective intervention strategies under operational and environmental uncertainty. We present a modeling and optimization framework for aerial wildfire suppression that combines a hybrid neural-cellular automaton wildfire model with gradient-based design of targeted aerial drops. The wildfire model predicts spatially varying spread behavior from terrain, fuel, and wind data, while the intervention module determines binary drop actions with continuous-valued location and orientation parameters mapped to the simulation grid. Water and retardant are represented with distinct suppression effects, corresponding to immediate reduction of active burning and persistent reduction of future spread. To evaluate the robustness of the resulting suppression plans, we quantify both aleatoric uncertainty through Monte Carlo sampling of daily fire-state realizations and epistemic uncertainty through spatially correlated prediction-error perturbations. A case study based on the 2020 Bear Fire shows that the framework can generate coherent aerial suppression schedules for reducing total fire-affected area and can support uncertainty-aware analysis of wildfire intervention strategies.

The choice of speech representation is critical in speech-driven 3D facial animation. Representations differ in what they encode: SSL features emphasize segmental and semantic cues, neural codecs yield latents optimized for acoustic reconstruction, and ASR-style objectives produce label-based spaces. We evaluate four speech representation families for 3D facial synthesis, comparing their facial reconstruction quality across two facial decoders using objective metrics and a perceptual evaluation. We additionally conduct probing analyses that relate tokenized representations to phonetic units and to articulatory deformations. We found that encoding phonetic classes is beneficial for accurate facial animation prediction on both semantic and label-based representations with comparable facial animation quality. From the latter, we introduce an Audio Visual Text-to-Speech (AVTTS) pipeline that leverages, as a shared space, discrete representations to decode speech and 3D facial motion.

There is a proliferation of work arguing for the use of synthetic data in scientific research. For example, social scientists are arguing for the use of LLM-generated "silicon samples" in pilot studies; AI evaluations increasingly rely on "LLM-as-a-judge" outputs; and proteomics research is accelerated by generative models that produce synthetic protein structures. These developments raise an intriguing possibility: synthetic data may help researchers ask more questions, run more studies, and accelerate discovery. But they also raise a fundamental concern: synthetic data can be biased, noisy, and misspecified. In this work, we propose statistical principles for using synthetic data in scientific research with provable validity guarantees. The key insight is a new technical condition that we call task exchangeability. Informally, this is a requirement that the researcher can identify historical tasks, for which real data is available, such that their current task of interest is exchangeable with the historical tasks in an appropriate mathematical sense. We develop methods for valid inference under task exchangeability, together with extensions that provide guarantees even beyond exchangeability. We demonstrate the framework on public opinion surveys with silicon samples and AI evaluation with autoraters.

We study generative modeling of Bach-style symbolic piano music using a shared MIDI corpus and three model families: autoregressive LSTMs with attention, latent-variable models including recurrent VAEs and vector-quantized VAEs, and generative adversarial networks. We compare their ability to model polyphonic note sequences, learn useful latent representations, and generate stylistically coherent compositions. Our experiments show that the autoregressive LSTM with attention produces the most musically coherent samples, while vector quantization helps mitigate posterior collapse and yields more structured outputs than conventional recurrent VAEs. The adversarial approach captures local pitch patterns but remains difficult to train and generalizes less reliably to Bach's style. These results highlight the relative strengths and failure modes of autoregressive, latent-variable, and adversarial approaches for symbolic music generation.

Large language models (LLMs) have enabled time series (TS) analysis by jointly modeling numerical observations and textual context through a shared token interface. However, TS tokens and prompt tokens exhibit fundamentally different information structures, making uniform token processing inefficient. In this paper, we study token efficiency in TS language modeling from an asymmetric-token perspective. We show that TS tokens have highly uneven spectral contributions, where many tokens share redundant frequency patterns while a small subset preserves critical temporal evidence. We also observe that prompt-token influence attenuates with model depth, suggesting that full prompt retention across all layers is unnecessary. Based on these findings, we develop an adaptive token budgeting framework that compresses TS tokens via frequency-domain structure and progressively reduces prompt tokens across layers. Experiments across forecasting, classification, imputation, and anomaly detection demonstrate up to \textit{\textbf{7.68$\times$}} inference acceleration and performance gains in \textit{\textbf{78\%}} of evaluated settings, showing the effectiveness of asymmetric token compression for scalable TS foundation models.

As parallel workloads grow in complexity, managing fine-grained data dependencies becomes a critical challenge. Futures offer a promising model for handling these dependencies, particularly in irregular algorithms, but they also come with the restriction of value-immutability. This immutability limits the ability to perform in-place memory updates, a necessity for high-performance linear algebra where memory recycling is paramount. In this paper, we address these limitations by introducing a new construct, await_delete, which extends traditional future semantics to allow safe value reuse once consumers are finished. Building on this extension, we present a novel future-based algorithm for the block-wise inversion of dense, symmetric matrices, motivated by a recent algorithm for finding conservation laws of dynamical systems. We implement our approach in an extended version of Taskflow and evaluate it through strong-scaling experiments. Our results demonstrate that while futures incur significant overhead on smaller problem sizes, they achieve nearly linear scaling on large matrices. We analyze the amortization threshold and show that futures are a viable high-performance tool for large-scale linear algebra.

Shielded reinforcement learning is typically presented as a runtime safety mechanism that compiles temporal-logic specifications into automata restricting an agent's actions. We argue this is the wrong product. The same automata-theoretic machinery -- specification compilation, product game construction, attractor computation, and winning-region extraction -- is better read as a design-time analytical instrument whose outputs are structural insights about a system rather than runtime constraints on a deployed agent. We instantiate this through a constrained two-player safety game for network defense. The two specifications are enforced asymmetrically: the defender specification defines the unsafe region of the game, whereas the attacker specification restricts the adversary's legal actions during attractor computation. Solving the game yields a defensibility verdict -- a formal certificate that a topology-specification pair is or is not defensible -- with the associated winning region and shield. Beyond the binary verdict, we derive topology-level metrics from the attractor structure and combine them with post-convergence behavior from shield-constrained adversarial multi-agent reinforcement learning. Together these form a defensibility fingerprint capturing both a network's formal safety properties and its operational behavior under adaptive play. A what-if analysis shows that formal defensibility and operational effectiveness capture distinct aspects of security: small architectural changes can produce large shifts in operational outcomes while leaving formal safety margins nearly unchanged. Shield synthesis is thus most valuable not as a deployment mechanism for safe agents, but as a framework for answering architectural questions about whether, where, and how a system can be defended. The defensibility verdict is the output, not the safe policy.

Frontier coding agents now promise end-to-end authorship of complete software systems. Two empirical questions follow: can AI coding-agent teammates program in any target language, including ones with no comparable prior open-source artefact? If so, does language choice still shape the artefact, and along which dimensions? We study both through a polyglot case study built around chess engines: non-trivial multi-component systems that admit a hierarchy of language-agnostic oracles, from exact move-generation correctness to a strength scale (Elo), observable from Rust to Brainfuck. We prompted two frontier agents (Claude Code and Codex) at the capability level, without chess knowledge or implementation guidance, under a documented intervention and stopping policy. The agents produced 34 chess engines spanning 17 primary programming languages, from mainstream to specialised, domain-specific, legacy, and esoteric targets. We combine per-engine feature analysis, independent Elo assessment, and session trajectories with qualitative analysis of code and transcripts. Frontier coding agents are genuinely polyglot: every language we tried produced at least one feature-rich working engine, several with no prior open-source counterpart of comparable scope (e.g., LaTeX), and the code is synthesised from scratch rather than copied. Yet language choice still matters: strong playing strength is only reachable in mainstream compiled languages, cost and engineering effort grow sharply as the language becomes more exotic, and feature choices shift across language families. Agents validate their own work unprompted, but their strength self-estimates are biased and a few engines cheated by calling a chess library. Programming language is no longer about whether AI teammates can build a working system, but about performance, cost, what gets built, and how much human supervision validation still needs.

We give a short proof that the majority vote of three independent consistent classifiers is an optimal learner in the realizable PAC setting. This proves optimality for the simplest voting scheme, while simplifying both the algorithmic structure and the probabilistic analysis of previous voting learners, including the algorithm of S. Hanneke and the analysis of bagging by K. Green Larsen.

Search-augmented LLMs increasingly mediate everyday consumer recommendations by retrieving live web content. This creates a new risk: generative recommenders may consume polluted web content, such as fake reviews and promotional pages crafted to mislead recommendations. We ask: to what extent do search-augmented LLMs become unwitting promoters of fake products when consuming polluted retrieval results? To answer this, we introduce FORGE (Fake Online Recommendations in Generative Environments), a benchmark for measuring fake-product promotion under controlled web-content pollution. Given an upstream search result, FORGE locally rewrites real products in retrieved web pages into fake ones to simulate web-content pollution, and measures how often the LLM recommends the fake product. FORGE covers 225 real-world products across 15 categories and 5 consumer scenarios. Across 12 commercial and open-weights LLMs, all models are vulnerable: a single polluted page yields fooled rates of up to 27%, while the full top-3 replacement raises this to 73.8%. Vulnerability varies substantially across categories, increasing when models lack stable prior knowledge of the relevant products. Reasoning does not mitigate this vulnerability; instead, it often generates spurious social proof to justify false recommendations. We evaluate three defenses: skepticism prompting and consensus filtering (over model priors or cross-document evidence). Skepticism can exacerbate vulnerability, much like reasoning, while filtering risks suppressing legitimate products. We release FORGE at https://github.com/leoluolol/forge-benchmark.

Agent systems are advancing quickly across domains, but their evaluation remains fragmented. Most benchmarks rely on fixed, LLM-centric harnesses that require heavy integration, create test-production mismatch, and limit fair comparison across diverse agent designs. The root problem is the lack of an open, agent-agnostic assessment interface. We advocate Agentified Agent Assessment (AAA), where evaluation is performed by judge agents and all participants interact through standardized protocols: A2A for task management and MCP for tool access. Conventional benchmarking defines two separate interfaces, one for the benchmark and one for the agent, while AAA only needs one; this yields a generic, unified framework that separates assessment logic from agent implementation and enables reproducible, interoperable, and multi-agent evaluation. We further introduce AgentBeats as a concrete realization of AAA: we identify five practical operation modes that make standardized assessment compatible with real-world constraints on openness, privacy, and reproducibility. To evaluate our design at scale, we conduct two studies: a five-month open competition that drew 298 judge agents across 12 categories together with 467 subject agents from independent participants, showing that AAA applies across a heterogeneous range of benchmarks; and a case study on coding agents that confirms agentified evaluation preserves fidelity with the public record while surfacing previously missing head-to-head results, yielding research insights about agent design. Combining a community-scale field study and a controlled coding case study, we verify that AAA delivers coverage, practicality, and fidelity across heterogeneous scenarios at scale. Together, AAA and AgentBeats offer a clear path toward open, standardized, and reproducible agent assessment.

When large language models (LLMs) fail to generalize or make haphazard errors in reasoning, it is often taken as evidence that LLMs are not truly reasoning, but rather performing a kind of pattern matching. The implication is that people's behavior does not exhibit the same types of failures because human reasoning uses principled and abstract world models. We evaluate human participants and 25 LLMs on their ability to engage in common-sense reasoning about a variety of everyday situations and observe similar patterns of errors in both people and models. We then identify the set of attention heads driving LLM responses and find that these heads implement a form of pattern-matching. These attention heads allow us to predict seemingly inexplicable reasoning errors in people caused by ostensibly irrelevant prompt details. Taken together, our results suggest that everyday causal reasoning in people and LLMs is more consistent with a form of pattern-matching than with abstract world models.

This paper presents a distribution-agnostic robust trajectory-optimization framework based on chance-constrained reinforcement learning. The uncertainty is represented here through initial conditions and process noise, with the only requirement being that it can be sampled. A deterministic nominal trajectory is first computed offline, and reinforcement learning is then used only to robustify that baseline through a structured affine closed-loop correction law comprising a feedforward control adjustment and time-varying feedback gains. Probabilistic feasibility is enforced empirically through rollout-based upper-tail quantiles, while terminal dispersion is regulated through covariance-feasibility penalties. The framework is assessed on two materially different trajectory design problems. The flagship case study is a three-dimensional multi-impulse Earth-Mars transfer, where the learned policy is benchmarked against a recent robust trajectory-optimization reference under Gaussian uncertainty and then evaluated under bounded uniform uncertainty and under process disturbances not seen during training. The second case study is a stochastic atmospheric pinpoint rocket landing problem, used to assess portability to a short-horizon continuous-thrust setting with drag, mass depletion, and glide-slope constraints. The results show that the proposed framework can remain competitive in upper-tail fuel cost while preserving probabilistic feasibility, and that the same robustification scaffold can be carried across heterogeneous spacecraft trajectory planning problems without redesign of its core stochastic-control structure.

Dispatch in three-sided marketplaces provides a natural setting for reinforcement learning from world feedback: decisions are evaluated by delayed operational outcomes such as delivery speed, courier utilization, and merchant congestion. We present a deployed reinforcement learning system at DoorDash that adapts dispatch objective weights in a large-scale food-delivery marketplace using delayed signals. Rather than replacing the combinatorial assignment optimizer, a store-level policy learned from logged marketplace data selects a discrete multiplier that shifts the dispatch optimizer's tradeoff between delivery quality and batching efficiency. This interface enables offline policy learning under noisy, delayed, and coupled feedback while preserving production feasibility constraints and operational safeguards. We train a shared value function using centralized offline data and decentralized store-level execution, with Double Q-learning targets and a conservative regularizer to reduce out-of-distribution value overestimation. In a production switchback experiment, the offline-trained policy increases batching and reduces courier-side time costs without degrading customer-facing delivery quality. Results illustrate how world feedback from a live economic and logistics system can be used to safely adapt decision policies online.

Chain-of-thought (CoT) reasoning is the dominant paradigm for inference-time scaling in language models, yet the causal influence of individual steps on the final answer poorly understood. We estimate each step's causal importance via early exit and use this measure to study how answers form across the reasoning traces of several model families. Across diverse tasks, we find that reasoning typically crosses a \emph{commitment boundary} -- a sharp transition from transient intermediate guesses to a stable, high-confidence answer. This transition often happens in a single step, well before the model's reasoning block ends, and is followed by \emph{epiphenomenal} CoT steps that leave the final answer probability unaltered. Using attention probes, we show that answer-formation stages can be linearly decoded from intermediate reasoning steps with high accuracy and generalize robustly to unseen reasoning tasks. We exploit this signal to early-exit reasoning blocks at the commitment boundary, reducing the length of CoTs up to 55\% on average with negligible impact on model performance.

We introduce EpiBench, a verifiable benchmark for short-horizon epigenomics analysis. EpiBench evaluates whether agents can make well-defined analysis decisions from realistic workflow states and return deterministically gradable answers. The benchmark includes 106 evaluations across CUT\&Tag/CUT\&RUN, ATAC-seq, ChIP-seq, and DNA methylation workflows. Across 5,088 valid trajectories from 16 model-harness pairs, no system passed a majority of attempts: GPT-5.5 / Pi led at 45.0\% (143/318 attempts; 95\% confidence interval (CI), 36.3--53.7), followed by GPT-5.5 / OpenAI Codex at 39.9\% (127/318 attempts; 95\% CI, 31.6--48.3). Claude Opus 4.8 Max / Pi and GPT-5.4 / Pi each passed 39.0\% (124/318 attempts; 95\% CI, 30.2--47.8 and 31.0--47.0, respectively). Performance varies across assay types, and many failed runs still contain parts of the correct answer. Agents often found the right files and computed useful intermediate results, but failed when the task required deeper, assay-specific scientific judgment.

Multi-Agent Systems (MAS) built on Large Language Models (LLMs) require effective orchestration to coordinate specialized agents, yet training such orchestrators is hindered by limited supervision and high computational cost. We propose Orchestration Reward Modeling (OrchRM), a self-supervised framework for evaluating orchestration quality without human annotations. OrchRM leverages intermediate artifacts from multi-agent executions to construct win-lose pairs for Bradley-Terry reward model training. Unlike existing MAS test-time scaling and orchestrator training frameworks that rely on costly sub-agent rollouts, OrchRM operates directly at the orchestration level, enabling efficient and high-performing reward-guided orchestrator training and MAS test-time scaling. OrchRM improves training efficiency by up to 10x in token usage while improving MAS test-time scaling performance by up to 8% in accuracy. These gains consistently transfer across multiple domains, including mathematical reasoning, web-based question answering, and multi-hop reasoning, demonstrating orchestration-level reward modeling as a scalable direction for robust multi-agent orchestration. Code will be available at https://github.com/Wang-ML-Lab/OrchRM.

Multi-agent systems communicate mostly through text, paying a lossy and expensive decode and re-encode cost. KV-cache communication is a promising alternative, yet most prior work is homogeneous, using duplicate copies of the same model, and avoids the central challenge of cross-model latent alignment; existing heterogeneous methods are also restrictive, typically assuming shared input and using transferred caches mainly for steering. We study a more fundamental question: can heterogeneous agents be aligned well enough to perform real "mind reading" and transfer both what one agent sees and how it thinks? Our information-structure analysis reveals a duality: context-aware transfer is driven by sparse reasoning signals, while context-unaware transfer, where the receiver sees no input, requires dense contextual knowledge preservation. Motivated by this, we propose dense alignment for heterogeneous KV-cache communication via a lightweight cross-model cache transformation and two-phase training: reconstruction followed by generation. Across all six directions of {Qwen3-4B, 8B, 14B} and six in-domain and out-of-domain benchmarks, our method outperforms prior heterogeneous baselines, matches or exceeds text communication in context-aware settings at roughly 2 to 3 times lower compute, and remains effective in context-unaware transfer where prior methods collapse.

Confidence is used for reliability, oversight, and a range of downstream decision tasks in Natural Language Processing (NLP), yet no existing method produces or evaluates a confidence for the output of a multiagent system. Prior work uses confidence within multiagent debate (MAD) to weight messages, trigger debate, or calibrate individual agents, but it never aggregates these into a single confidence for the system itself. We introduce three protocols that produce a final answer along with a single aggregated confidence by first transforming raw confidence signals to make them comparable across models, then combining them via soft voting or a probability fusion we call Bayesian fusion. This aggregated confidence is substantially more discriminative (AUARC) than that of the best single agent or the standard debate baselines, while correctness (F1-score) stays stable and recovers the losses MAD incurs on more ambiguous tasks. Analyzing two estimators, sequence probability and self-report, alongside parametric and non-parametric calibrators, we find that calibration improves F1 for both estimators while AUARC is less reliant on it. We evaluate six homogeneous and heterogeneous debating pairs per benchmark, across five benchmarks and four task types, spanning a range of model capabilities and sizes.

We present Simplex-Constrained Sparse Bagging (SCSB), a mathematically rigorous framework for post-training compression and probability calibration of bootstrap-based bagging ensembles. Standard bagging ensembles (such as Random Forests, Bagged SVMs, and Bagged Neural Networks) assign uniform voting power to all constituent estimators. However, this naive uniform prior ignores the varying local competence of base estimators and contributes to model overconfidence. We formulate ensemble pruning and calibration as a joint optimization problem over the probability simplex by minimizing the Out-Of-Bag (OOB) loss. To induce sparsity, we address the theoretical "L1-simplex paradox" -- the mathematical reality that the L1 norm is constant on the simplex and fails to prune -- by introducing a concave quadratic penalty. SCSB is model-agnostic and achieves up to 96% ensemble compression, yielding linear inference speedups and superior probability calibration (lowered Expected Calibration Error) while preserving or enhancing generalization accuracy.

Despite raising concerns about the mental health effects associated with the usage of TikTok, little is known about how related content is framed by creators and received by audiences. We collect the content of 28,341 TikTok videos and 80,130 comments from Mental Health Awareness Month (May) in 2023 and 2024 via the TikTok Research API, and study how the tone of awareness varies across topics and years. We characterize "tone" as the emotional and interpersonal framing of mental health discourse, operationalized through sentiment and toxicity measures. We extract topics from video text using BERTopic and log-odds keywords, then quantify topic-conditioned sentiment (XLM-T) and toxicity (Detoxify) separately for video transcriptions and comments. Sentiment captures the affective valence of content, while toxicity reflects the presence of harmful or abusive language. We find a stable set of recurring themes across years, spanning clinical conditions, emotional disclosure, self-care, and campaign-oriented content, with engagement highly skewed toward a small subset of topics. All sentiment and toxicity analyses are computed separately for video content and comments, allowing us to distinguish between content production and audience reception. Sentiment in videos is often negative for emotionally charged topics, while comments tend to shift toward more mixed or positive polarity, especially for suicide prevention. Toxicity is low in median overall, but exhibits longer-tailed outliers in comments than in videos that are more pronounced in comments and concentrated in specific topics (e.g., "Duet", "Suicide Prevention", and "Psychisch"). Overall, our results provide a topic-level decomposition of mental health discourse on TikTok during awareness-month campaigns.

Event-based vision has drawn increasing attention owing to its distinctive properties, including ultra-high temporal resolution and extreme dynamic range. Recent works have introduced it to video super-resolution (VSR) to enhance flow estimation and temporal alignment. In contrast, this paper shifts the focus of event signals from motion refinement to texture enhancement in VSR. We propose EvTexture++, the first event-driven framework dedicated to texture enhancement in VSR. It leverages high-frequency spatiotemporal details from events to improve texture recovery. EvTexture++ incorporates a customized texture enhancement branch, along with an iterative texture enhancement module that progressively exploits high-temporal-resolution event information for texture restoration. This enables gradual refinement of texture regions across iterations, yielding more accurate and detailed high-resolution outputs. Besides intra-frame texture recovery, large motions could degrade inter-frame temporal consistency, particularly in texture regions, leading to texture flickering. To mitigate this, we further exploit the continuous-time motion cues of events to enhance temporal consistency, introducing a temporal texture alignment module that estimates event-guided texture-aware flow for precise inter-frame texture alignment. Moreover, EvTexture++ is designed as a plug-and-play tool to flexibly boost the performance of existing VSR models. Experiments on five datasets demonstrate that EvTexture++ achieves state-of-the-art performance. When integrated into recent VSR models, it yields significant improvements, with gains of up to 1.55 dB in PSNR on the texture-rich Vid4 dataset. Code: https://github.com/DachunKai/EvTexture.

Scientific laboratories increasingly rely on AI systems to reason about experiments, but the physical act of doing science remains largely outside their reach. AI can help read literature, generate hypotheses, and plan protocols, yet the execution of those protocols at the bench still requires a human operator. Vision-Language-Action (VLA) models provide one possible interface between written protocols and robot execution, but existing policies are trained mostly on household and tabletop demonstrations and rarely encounter the instruments, transparent liquids, or fixed protocol workflows found in scientific laboratories. Closing this gap requires both laboratory-specific supervision and a unified learning framework that can accommodate the diverse robot embodiments used to execute experimental protocols. We therefore identify data and embodiment as central bottlenecks alongside model design. To address the data side, we build RoboGenesis, a simulation-based workflow and data engine that composes configured laboratory workflows from atomic skills, validates and filters rollouts, and exports structured demonstrations across supported robot profiles. On the policy side, we present LabVLA, trained with a two-stage recipe: FAST action token pretraining first makes the Qwen3-VL-4B-Instruct backbone action aware before any continuous control is learned, and flow matching posttraining then attaches a DiT action expert under knowledge insulation. On the LabUtopia benchmark, LabVLA achieves the highest average success rate among all evaluated baselines under both in-distribution and out-of-distribution settings.

Understanding the minimal assumptions necessary for generalization is the fundamental question in learning theory. Unfortunately, most results rely heavily on independence (or some proxy thereof) of the data-generating process, while results for strongly dependent data are far more limited. Towards addressing this gap, we introduce the framework of simulatable processes, where the learner has access to a simulator that approximates the distribution generating the data (which may be an arbitrarily complex and dependent process). Surprisingly, given access to such a simulator, we show that we can recover the same learning guarantees as in the classical setting with independent data, namely, error bounds that depend on the VC dimension. Further, we use this framework to study the power of conditional sampling and show strict statistical and computational advantages in this setting. As a highlight of our framework, we exhibit a single algorithm that simultaneously learns any given VC class under all processes samplable in bounded polynomial time, with regret controlled by the time-bounded Kolmogorov complexity of the process. This provides a significant conceptual broadening of the classical PAC model.

Multimodal Large Language Models (MLLMs) have shown promising reasoning capabilities in general domains, yet their performance remains limited in specialized settings such as healthcare, especially in multilingual and low-resource scenarios. This gap is critical in regions like rural India, where patients often express complex medical queries in native Indic languages and rely on multimodal inputs such as medical images. Existing English-centric MLLMs struggle to support such use cases, limiting equitable access to AI-driven healthcare assistance. To address this challenge, we introduce ArogyaBodha, a large-scale multilingual multimodal medical question-answer dataset constructed from eight heterogeneous sources, covering 31 body systems, six imaging modalities, and 21 clinical domains across English and seven major Indian languages. We further propose ArogyaSutra, an actor-critic-based multi-agent framework that integrates tool grounding with dual-memory mechanisms for step-wise, reasoning-aware decision making, and uses stored actor-critic simulation trajectories for distillation. Experiments show that our dataset and framework improve multilingual medical reasoning accuracy across all Indic languages, with ablations validating the contribution of each component. The source code and dataset are available at: https://iitp-cse.github.io/ ArogyaSutra/

Real-world time series are often highly incomplete and irregular due to sensor dormancy, transmission delays, and event-driven sampling, making reliable forecasting fundamentally challenging. Existing methods have evolved from impute-then-forecast pipelines to continuous-time models such as Neural ODEs and continuous-time graph networks. While these approaches improve the modeling of historical irregularity, they still rely on an implicit oracle assumption at inference time: the timestamps of future valid observations are presumed to be known in advance. This assumption limits practical relevance, since in many real systems the more fundamental question is not only what the future value will be, but also whether a valid observation will occur at all. In this paper, we propose Timeflies, a unified framework that reformulates forecasting as a joint problem of future observability inference and value estimation. To explicitly model the interaction between observation dynamics and state evolution, Timeflies adopts an observation stream and a value stream, coupled through three dedicated modules for reliability-aware embedding, observation-guided dependency modeling, and joint prediction. We further construct Shadow, a benchmark that combines natural missingness from public datasets with real-world industrial data, and introduce the Observation-Value Joint Entropy (OVJE) metric to comprehensively evaluate this coupled predictability. Extensive experiments show that Timeflies consistently outperforms existing methods, highlighting the importance of explicitly modeling future observability in time series forecasting with missing values. Code and dataset are available in https://github.com/ant-intl/Timeflies.

Many important observables in physics and geometry are cup products of cochains. The adjusted cup product neural layer has been introduced in this paper. It is a neural primitive that hard wires the cup product with an adjustment term from higher gauge theory. This creates a readout that is gauge invariant by design. Their main theoretical result shows that on a closed cycle the output relies entirely on the adjustment coefficient. Setting this coefficient to zero removes the output completely regardless of other parameters. Thus the adjustment is the only source of gauge invariant signal. They prove this observable is a nonzero quadratic form and is exactly invariant under one and two gauge transformations.

Current discussions of AI in scientific discovery are often dominated by two visible capabilities: search over existing knowledge and execution through optimization, simulation, and automation. Both are important, but neither fully captures the central act of discovery: the formation and evolution of models. This paper proposes a three-layer view of AI in discovery. Layer 1 is search and retrieval by large language models. Layer 2, as the main innovation of this paper, is model formation through qualitative reasoning: the capacity to recognize when a current framework is structurally inadequate and to understand the problem within a broader representational space, not through trial and error, but through structural insight into what is missing and where it can be found. Layer 3 is execution, optimization, and refinement. The main claim is that Layer 2 is both the most important and the least developed. Search without model formation remains confined to inherited frameworks, while execution without conceptual revision only amplifies an existing formulation. We illustrate Layer 2 reasoning through three case studies: S. S. Chern's intrinsic proof of the Gauss-Bonnet theorem, the resolution of the Nesterov Accelerated Gradient convergence problem via Lyapunov functions, and the autonomous disproof of the Erdos unit distance conjecture by OpenAI in 2026. Each case exhibits the same structural signature: a framework that had become inadequate, a missing conceptual object, and a resolution found in an unexpected neighboring field.

Discrete diffusion models offer a simple and stable likelihood-based framework for sequence generation, recently extended to any-length settings via token insertion. Principled reward-guided fine-tuning for any-length discrete diffusion, however, remains largely unexplored. We introduce Fine-Tuning Any-Length Discrete Diffusion for Adaptive Decoding (A2D2), a unified framework for reward-guided fine-tuning of any-length discrete diffusion models via joint optimization of the insertion and unmasking policies together with a quality-based inference schedule. We derive the Radon-Nikodym derivative for the joint insertion-unmasking path measures, enabling theoretically guaranteed convergence to the intractable reward-tilted sequence distribution without requiring target samples. Building on this, we establish unmasking and insertion quality as tractable approaches for minimizing decoding error and introduce the Adaptive Joint Decoding (AJD) loss, which provably yields the optimal path measure that generates the reward-tilted distribution. Empirically, A2D2 improves reward optimization while enhancing generation flexibility and accuracy over prior fixed-length fine-tuning and inference-time guidance methods.

Purpose: To investigate whether contrast-informed data augmentation and domain-adversarial training improve the adult-to-neonatal generalization of the E2E-VarNet. Methods: Three training regimes were investigated: (1) adult-only training with unaugmented adult data, (2) mixed training with paired unaugmented and neonatal-informed augmented adult data, and (3) mixed training with a domain-adversarial objective. Models were trained on retrospectively undersampled multi-coil adult T2-weighted brain MR data and evaluated on neonatal and adult test data at acceleration factors $R=4$ and $R=8$ using quantitative metrics and qualitative evaluation. Feature analyses assessed whether domain-adversarial training altered the latent representations of unaugmented adult, augmented adult, and neonatal test samples. Results: Mixed training (Mixed) and mixed domain-adversarial training (Mixed-DAT) outperformed unaugmented adult-only training (Unaug-Only) when evaluated on neonatal data. At R=4, Mixed-DAT achieved the best performance (SSIM = 0.924 +/- 0.027, PSNR = 33.98 +/- 1.15 dB). At R=8, Mixed-DAT performed best when measured using SSIM (0.848 +/- 0.031 vs. 0.766 +/- 0.037 for Unaug-Only and 0.814 +/- 0.035 for Mixed) and Mixed performed best when measured using PSNR (29.56 +/- 0.83 dB vs. 26.26 +/- 0.78 dB for Unaug-Only and 29.43 +/- 0.83 dB for Mixed-DAT). Qualitative assessment of t-SNE plots suggested that Mixed-DAT increased the overlap among the latent representations of the unaugmented adult, augmented adult, and neonatal test data. Conclusion: Contrast-informed augmentation and domain-adversarial training improved adult-to-neonatal generalization of deep learning-based MR reconstruction. These findings suggest that contrast-informed data augmentation combined with adversarial training may improve robustness to domain shift in undersampled neonatal MR reconstruction.

Spiking Neural Networks (SNNs) provide an attractive framework for energy-efficient inference due to their event-driven computation and biologically inspired dynamics. However, efficient hardware realization of SNNs remains challenging because neuronal computations incur significant power and area costs, and uncontrolled approximate arithmetic can destabilize recurrent state updates when precision is not properly managed. To address these challenges, this paper presents ReSCom, a reconfigurable SNN accelerator that leverages stochastic computing to reduce hardware complexity while maintaining stable inference. The proposed architecture employs stochastic arithmetic for multiplication operations in neuron dynamics, while preserving exact fixed-point addition/subtraction operations. This stochastic strategy enables runtime trade-offs between accuracy, latency, and energy consumption. A unified reconfigurable neuron design supports Integrate-and-Fire (IF), Leaky Integrate-and-Fire (LIF), and Synaptic neuron models within a single hardware framework. Experimental results for MNIST inference on a Xilinx Artix-7 FPGA show that ReSCom achieves $92.80\%$ classification accuracy while consuming just $0.05~\mathrm{mJ}$ of operational energy per image at $100~\mathrm{MHz}$, outperforming the energy efficiency of recent state-of-the-art implementations. Furthermore, managing the stochastic bit-stream length allows explicit, dynamic control over accuracy-latency-energy trade-offs to meet target application constraints.

Text-guided image editing with visual autoregressive (VAR) generators requires controlling both what the model samples and where the sampled change is written back into the image code. Existing VAR editors mainly operate on token streams, features, or flat next-token logits, leaving two native structures of bitwise-residual VAR models underused: the per-bit Bernoulli prediction head and the additive multi-scale residual code field from which the image is assembled. We propose BitResEdit, a training-free editor for bitwise-residual VAR generators such as Infinity. BitEdit performs source-negative guidance by tilting the post-CFG per-bit log-odds along a source--target contrast computed on a shared edited prefix, then projects each update into a closed-form Bernoulli-KL trust region around the clean CFG sampler. ResEdit converts the sampled bits into per-scale continuous-code residuals, gates them with a localization mask, and re-injects them through the generator's native sum-of-scales. Together they couple decision-time bit guidance with combination-time code composition, so masked-out latent features are preserved exactly by code arithmetic while localized, scale-aware edits are applied inside the target region. On PIE-Bench with Infinity-2B, BitResEdit attains the strongest text alignment among same-backbone VAR editors, improving CLIP on the edited region by +1.07 over the strongest prior editor while keeping background preservation competitive with it. Ablations show BitEdit and ResEdit play complementary roles in target alignment and background preservation.

Personalized health AI systems face a fundamental cold-start problem: machine learning models for physiological interpretation require weeks of individual behavioral data before they can distinguish constitutional variation from environmentally driven deviation. We propose a solution grounded in causal inference and Bayesian prior design. An individual's genomic profile serves as an exogenous genetic anchor -- a domain-informed, personalized prior that is fixed at conception, immune to reverse causation, and available before a single behavioral observation is collected. The anchor initializes a Bayesian belief state over an individual's physiological set point G-hat = mu + sum(beta_i * g_i), where beta_i are GWAS-derived effect sizes and g_i are risk-allele counts. Each incoming physiological measurement P produces a non-constitutional deviation delta = P - G-hat that separates the signal attributable to environment and state from the constitutionally fixed baseline. As behavioral data accrue, the prior decays according to G-hat_t = w(t)*G-hat_genomic + [1-w(t)]*P-bar_t, transitioning from genome-dominated to empirical-baseline-dominated inference. The same observed HRV of 55 ms generates a suppression hypothesis for a person whose prior predicts 80 ms, and an enhancement hypothesis for a person whose prior predicts 30 ms -- a reversal impossible without a personalized anchor. We develop this architecture across six physiological domains, grading genomic priors by evidence strength, distinguishing robustly replicated anchors (FTO, FADS1/2, FKBP5) from contested candidate genes (SLC6A4, MAOA, DRD2). We address the inference boundary between association, Mendelian randomization, and individual token causation, and define four constraints for deployment: evidence-graded priors, dynamic decay, ancestry-matched effect sizes, and attribution rather than deterministic output.

Large language model (LLM) reviewers are increasingly used in pull-request (PR) workflows, where their approvals help decide which code is merged into a repository. This raises a question that benchmarks for static vulnerability detection or code generation do not address: can an automated reviewer reject a malicious contribution when the attacker controls both the code change and the accompanying PR text? We introduce SEVRA-BENCH (Social Engineering of Vulnerabilities in Review Agents), a benchmark that measures how often an automated reviewer approves such adversarial pull requests. Each malicious PR in SEVRA-BENCH is built from a real project commit that previously fixed a vulnerability listed in the Common Vulnerabilities and Exposures (CVE) database. We automatically invert that fix to restore the original vulnerable code and submit it as a pull request wrapped in one of 15 social-engineering framings, which vary the claims made, the supporting evidence, the urgency conveyed, signals of prior approval, and appeals to authority. SEVRA-BENCH contains 1,062 malicious PRs drawn from Common Vulnerabilities and Exposures (CVE)-linked fixes across the top 10 entries of the 2025 Common Weakness Enumeration (CWE) Top 25. In a realistic setting, we evaluate 8 current LLMs as code review agents on PRs that introduce vulnerabilities previously reported in public disclosures. Our results reveal a sharp gap in security capabilities between closed- and open-source models. We hope SEVRA-BENCH will serve as a valuable resource for advancing open-source models and narrowing this gap.

Retrieval augmented generation (RAG) depends critically on the quality and granularity of retrieved evidence. Large retrieval units preserve context but often introduce irrelevant content, which can dilute answer bearing evidence and worsen long context utilization. Fine-grained units are more compact, but they may be difficult to retrieve reliably because short chunks can lack semantic, lexical, or bridging cues needed to match the query. We propose Uncertainty-aware Multi-Granularity RAG (UMG-RAG), a training-free hybrid retrieval framework that treats chunk granularity as query-specific reliability estimation. Instead of training a new retriever or modifying the generator, UMG-RAG uses existing dense and sparse retrievers as complementary experts across multiple chunk granularities. For each query, it converts each expert-granularity score list into an evidence distribution, estimates reliability from distribution entropy, and fuses candidates according to query-specific semantic, lexical, and granularity confidence. We further introduce UMGP-RAG, a parent promotion variant that uses fine-grained hits to locate relevant evidence while returning broader non-redundant parent chunks for local coherence. Experiments on question answering benchmarks show that uncertainty-aware fusion and parent promotion improve generation quality while maintaining a lightweight, plug-and-play retrieval pipeline.

Turn-taking in multi-party spoken conversations remains a fundamental challenge for voice-based agents, particularly under dynamic floor competition and varying user expectations. We propose ModeratorLM, a role-playing voice agent that conditions turn-taking behavior on an explicitly assigned role in multi-party settings. The system is built on a speech large language model operating in chunk-wise streaming manner. We further introduce a reasoning-augmented variant that incorporates chain-of-thought reasoning over conversational context and the assigned role. We construct RolePlayConv, a large-scale synthetic dataset of spoken multi-party conversations with diverse assistant roles. Experiments on real-world meeting data and RolePlayConv show improved turn-taking precision by over 40% and recall by more than 70%, while substantially reducing false-positive interruptions compared to non-role-conditioned baselines.

Can a learned model capture how faults propagate through a large-scale network and use this knowledge to causally attribute customer impact to its underlying root cause? Existing root cause analysis techniques often rely on static rules, correlation heuristics, or topology-local reasoning, which struggle to generalize in dynamic environments where faults propagate across complex physical and logical dependencies. We present NetCause, a self-supervised learning-based framework that models network incidents as graph-temporal processes and uses counterfactual simulation to rank candidate root causes. This approach produces an interpretable ranking of root cause hypotheses and integrates naturally with operator-defined mitigation and remediation actions. We train the model on over 1,500 incidents collected over six months from a leading cloud provider's production network and evaluate it on 31 expert-labeled incidents. NetCause consistently improves root cause ranking quality in the regime most relevant to operational decision-making, achieving a 16.1% accuracy improvement over a rule-based heuristic baseline. While training is computationally intensive, inference is lightweight, requiring only seconds of GPU runtime per incident (well below typical telemetry collection latencies).

While mixed-language querying is ubiquitous in multilingual communities, the sensitivity of dense retrievers to such queries remains poorly understood. We present a ratio-controlled study on mMARCO that systematically evaluates retrieval performance by varying the mixing proportion of parallel query translations via embedding-level mixing -- constructing mixed queries as an interpolation of monolingual embeddings. Experiments with BGE-M3 demonstrate that an optimal mixing ratio outperforms the best monolingual endpoint in 88/105 cases. We uncover a distinct asymmetry driven by English dominance: mixing is uniformly beneficial when retrieving from non-English document indices, whereas indices containing English are best served by pure English queries. Furthermore, English acts as the strongest mixing partner for every non-English document language. Finally, when controlling for English dominance, mixing gains correlate negatively with typological distance. We conclude that language-mix sensitivity is structured and predictable, and we validate the robustness of these patterns across model families and scales.

Cloud-computing relies on large-scale networks which are inherently complex systems. In this paper, we present a novel approach to root cause analysis (RCA) of cloud network incidents, leveraging graph-based causal discovery techniques. Our method addresses the limitations of rule-based automation by introducing a spatiotemporal grouping strategy and an automation ontology to reduce the dimensionality of the problem. We construct a causal graph from binary time series data using bivariate Granger causality and conditional independence tests. For inference, we introduce a probabilistic method that assigns edge-specific conditional probabilities as a function of time lag, allowing for interpretable, time-aware root cause scoring via causal graph traversal. We evaluated the system using a labeled dataset of 35 production incidents from a major cloud provider. The model successfully recalled the correct root cause in 85.7% of incidents and produced an exact match in 74.3%. In production, the deployed system has been used in over 800 real-world incidents, with positive qualitative feedback from network engineers. These results highlight the practicality of a data-driven, causal approach to RCA in dynamic and large-scale operational environments.

Post-traumatic stress disorder (PTSD) in veterans is characterized by persistent hyperarousal and comorbid anxiety and depressive symptoms that are difficult to monitor and manage outside clinical settings. Thirteen veterans participating in a Project Hero cycling event in Texas were randomized by computer-generated sequence in a naturalistic setting to two arms: (1) digital intervention plus physical activity, or (2) physical activity only, plus a third at-home monitoring control cohort consisting of 7 veterans selected from the broader Project Hero veteran community. Continuous smartwatch sensing combined heart rate and accelerometer features to detect hyperarousal events, which were confirmed in real time by participants. Weekly self-report measures of anxiety, depression, and PTSD severity were collected. Generalized additive mixed models characterized nonlinear trajectories over time. Baseline-normalized hyperarousal trajectories differed significantly across conditions, with the digital intervention group (n=7) showing structured stabilization compared to late-study escalation in the physical-only group (n=3). Both cycling groups exhibited acute symptom improvements during the endurance event; however, the digital intervention group demonstrated a higher overall maintenance of gains. The at-home control group (n=4) showed gradual symptom declines. Perceived precision of ML detections varied substantially across individuals and was positively associated with symptom severity, with higher-severity participants confirming a greater proportion of detected events. These results suggest that coupling wearable detection with digital self-management tools may support stabilization of hyperarousal and symptom improvement while emphasizing the importance of personalization and human-centered design in wearable mental health systems.

This paper presents a comprehensive overview of the QIAS 2026 shared task, organized as part of the OSACT7 Workshop and co-located with LREC 2026. The shared task was designed to evaluate the ability of large language models to perform complex reasoning in the religious and legal domain of Islamic inheritance. Unlike conventional question-answering benchmarks, QIAS 2026 focuses on end-to-end reasoning from natural language cases, requiring systems to perform the full inheritance calculation process, from identifying the eligible heirs to assigning the correct share to each beneficiary. To support this evaluation, the task was based on the MAWARITH benchmark, a dataset of $12{,}500$ Arabic inheritance cases annotated with intermediate reasoning steps and final answers. System submissions were evaluated using MIR-E, a multi-step metric that measures performance across the main stages of inheritance reasoning. A total of $16$ teams participated in the shared task, investigating a range of approaches, including prompting-based methods, retrieval-augmented generation, and fine-tuning strategies. The results show that Islamic inheritance remains a highly challenging benchmark for current language models, especially in stages that require precise legal interpretation and structured numerical reasoning. This overview summarizes the task design, dataset, evaluation framework, participating systems, and main results.

Frequency Modulated Continuous Wave (FMCW) radar systems traditionally rely on Fourier-based methods, such as the Fast Fourier Transform (FFT), to estimate target range and velocity. While computationally efficient, these approaches require storing and processing large blocks of data, which can become a bottleneck in memory-constrained or low-latency applications. In this work, we propose a neuromorphic-inspired signal processing method based on adaptive resonate-and-fire (ARF) neurons formulated as a discrete-time dynamical system. Each neuron dynamically adjusts its internal frequency to match dominant frequency components of the input radar signal, enabling direct estimation of target ranges and velocities without computing the full frequency spectrum. The proposed model operates in a sample-by-sample fashion, resulting in memory requirements that scale with the number of tracked targets rather than the signal length. A feedback mechanism is also introduced to enable multiple neurons to lock on distinct frequency components in multi-target cases. Results on simulated and experimental data demonstrate that the method can successfully track multiple targets. Compared to conventional FFT-based approaches, the proposed method offers reduced memory usage proportional only to the number of tracked targets, making it suitable for resource-constrained and edge-based radar applications.

We argue that aligning AI to aggregated human preferences is the wrong target. With current technology, one can train AIs to share the values of a Silicon Valley techno-optimist, a degrowth environmentalist, a national-conservative culture warrior, a single-party state cadre, or a devout religious traditionalist. We should not. Human values produce societies that thrive or fail on the merits of those values - from failed states and extreme inequality to declining happiness, political polarization, and government dysfunction in the world's wealthiest democracies. The pluralistic-alignment program correctly diagnoses that there is no single "humanity" to align with, but is dangerous if taken as the main directive. We argue that AI should be trained to a non-negotiable floor of objective alignment goals - competence, bounded by the constraints of factual accuracy, honesty, and lawfulness and that pluralism belongs at the surface (language, register, conventions, missing-context defaults) and across the wide band of legitimate value tradeoffs that respect the floor, but not at the level of values that violate it. We highlight the empirical reality of unfiltered pluralistic values, propose four commitments as a constructive alternative, and engage six credible objections: commercial pressure and practical feasibility, democratic legitimacy, regulatory compliance, over-reliance on institutionalist explanations, the charge that the floor itself is culturally laden, and the limits of Coherent Extrapolated Volition.

World Action Models (WAMs) present a promising paradigm for robotic control via video prediction. However, current WAMs suffer from fundamental spatial bottlenecks: standard text inputs introduce referential ambiguity in cluttered scenes, while unstructured RGB predictions lack semantic grounding and remain biased by task-irrelevant backgrounds. To overcome these limitations, we introduce MaskWAM, an object-centric world-action model. By jointly integrating masks as both explicit inputs and predictions via a unified Mixture of Transformers (MoT), MaskWAM unlocks robust policy generalization. This design provides two key benefits: (1) predicting future masks yields object-centric semantic supervision that suppresses visual noise, significantly enhancing even standard text-conditioned WAMs; and (2) coupling this predictive supervision with first-frame visual prompts, such as target object masks, establishes a precise spatial anchor that substantially reduces language ambiguity. Crucially, as WAMs are inherently vision-driven architectures, direct mask conditioning yields substantially stronger guidance than text alone, establishing a precise and robust paradigm for manipulating unseen objects. Evaluations on LIBERO, RoboTwin, and real-world tasks demonstrate that MaskWAM significantly outperforms baselines in both language-clear and language-ambiguous tasks.

Driven by conservative over-provisioning to guarantee service reliability, resource utilization in cloud data centers remains at low levels. To mitigate this, the forecast-then-optimize paradigm has emerged to optimize consolidation by anticipating future demands. While emerging time series foundation models promise to enhance this paradigm through zero-shot generalization, existing benchmarks focus solely on prediction error metrics. The actual decision utility of these advanced models remains unverified, rendering their practical value for downstream tasks uncertain. To bridge this gap, we propose CloudCons, a comprehensive end-to-end benchmark designed to evaluate forecasting models within the specific context of cloud resource consolidation. We build high-quality datasets that cover diverse workloads from Huawei Cloud, Microsoft Azure, and Google Borg, capturing distinct service characteristics ranging from synchronized diurnal rhythms to stochastic, pulse-like bursts and high-frequency noise. We conduct an extensive evaluation of statistical, deep learning, and foundation models. Our experiments reveal a pivotal finding: while foundation models demonstrate superior zero-shot forecasting accuracy, this advantage does not inherently translate into better decision utility. Of practical significance, we systematically analyze how the selection of predictive quantiles acts as a critical lever. We provide actionable guidelines for calibrating these selections to balance the trade-off between resource efficiency and service reliability, offering vital insights for real-world deployment decisions.

Anomaly detection is a fundamental component of intelligent systems with applications in healthcare, cybersecurity, smart grids, and IoT environments. Although conventional machine learning and deep learning methods have demonstrated effectiveness in identifying anomalies, they often rely on large labeled datasets, incur high computational costs, and face scalability challenges in edge and high-dimensional settings. This paper presents D2H-AD, a novel anomaly detection framework based on Hyperdimensional Computing (HDC), a brain-inspired paradigm that represents information using high-dimensional distributed vectors. Unlike existing HDC-based methods, D2H-AD integrates distance-based similarity and density-aware encoding within a unified framework, improving anomaly representation and detection performance. Ablation studies show that hyperdimensional encoding alone yields up to 5.4% higher ROC-AUC than applying the same density-distance scoring directly in the original feature space. Furthermore, D2H-AD consistently outperforms five established baselines, namely HDAD, ODHD, One-Class SVM, Isolation Forest, and Autoencoders, across all evaluated datasets. The framework is lightweight, interpretable, and computationally efficient, making it suitable for resource-constrained and real-time applications. We validate D2H-AD on five benchmark datasets and demonstrate superior F1-score and ROC-AUC performance, together with robustness to class imbalance, noise, and data complexity. In addition to improved accuracy, D2H-AD offers scalability, a small memory footprint, and low-latency operation enabled by binary computations and a compact design. These properties make it particularly attractive for TinyML and edge AI deployments. The proposed framework highlights the potential of HDC for accurate, interpretable, and energy-efficient anomaly detection in dynamic environments.

Indoor vision-based localization systems are affected by detection noise, occlusions, and limited camera coverage, leading to uncertainty at multiple stages of the pipeline. While multi-camera data fusion is widely used to mitigate these issues, it is typically treated as a black-box component and evaluated solely end-to-end, obscuring its mechanistic contributions. To address this gap, this work investigates whether explicitly characterizing single-camera localization errors can be leveraged to calibrate and optimize multi-camera data fusion. We introduce a measurement-calibrated fusion approach that integrates component-wise error quantification, specifically isolating homography calibration, human detection, and motion tracking. A component-wise evaluation is conducted to quantify error contributions from homography calibration, human detection, and motion tracking. Experimental results show that data fusion improves localization accuracy compared to single-camera baselines. While measurement-calibrated fusion provides only limited improvement in absolute accuracy over standard fusion, it substantially reduces trajectory variance and improves motion smoothness, which are critical for applications requiring stable and continuous motion estimates. These results highlight the value of explicit error characterization when designing data fusion strategies for vision-based indoor positioning systems.

Large-scale mined corpora provide abundant training data for end-to-end speech-to-speech translation (S2ST) but may contain noise, misalignment, and semantic errors. Filtering noisy data is crucial to maintain robust speech translation performance. We study how to train an audio-language model to make keep/drop decisions on paired speech directly from audio. To obtain reliable supervision without manual labels, we adopt a scalable two-stage Rank-to-Distill strategy. A lightweight ranker generates keep/drop pseudo-labels from noisy speech pairs, then trains an audio large language model to predict keep/drop directly from raw paired speech. The resulting model jointly captures acoustic fidelity and cross-lingual semantic consistency for the selection of speech-conditioned data. Experiments on CVSS-C and SpeechMatrix show consistent improvements over unfiltered training, yielding up to +1.4 ASR-BLEU for end-to-end S2ST.

Robust localization in unstructured environments, such as agricultural fields, is a critical challenge for autonomous systems. LiDAR sensors provide detailed 3D information about the environment and are invariant to lighting conditions. For this reason, LiDAR-based place recognition methods have gained significant attention. In this paper, we propose MinkUNeXt-VINE++, a novel approach that combines early fusion of heterogeneous LiDAR data from two sensors (Livox Mid-360 and Velodyne VLP-16) and a learned re-ranking strategy in inference time. This fusion leverages the strengths of each sensor to provide a more comprehensive representation of the environment. Additionally, the re-ranking approach is particularly important in repetitive environments, such as vineyards, as finding true positives is a major challenge. We evaluated our approach using the TEMPO-VINE dataset, which provides heterogeneous LiDAR data in vineyard environments across different phenological stages. Our results demonstrate that MinkUNeXt-VINE++ significantly improves place recognition performance compared to single-sensor approaches and state-of-the-art methods. MinkUNeXt-VINE++ achieves a 20% improvement in the Recall@1 metric compared to single-sensor approaches, and +30% including re-ranking. The code of our method is publicly available for reproduction.

Diffusion Transformers (DiTs) have become the dominant architecture for image and video generation, creating growing demand for efficient DiT serving. Existing systems assign each request a fixed parallel configuration throughout its lifetime. However, DiT workloads exhibit substantial heterogeneity across requests, execution stages, and system conditions, making static parallelism inefficient and often leading to poor GPU utilization and degraded service quality. This paper argues that DiT serving should treat GPU parallelism as a first-class schedulable resource. We present GF-DiT, a policy-programmable runtime for elastic DiT serving that dynamically adapts the parallelism of running requests according to workload demands and service objectives. GF-DiT introduces an asynchronous execution abstraction that decomposes requests into independently schedulable trajectory tasks and enables online GPU reallocation. To make elastic parallelism practical, GF-DiT further proposes group-free collectives, a lightweight communication abstraction that supports low-overhead online formation and reconfiguration of arbitrary execution groups. We implement GF-DiT in vLLM-Omni and evaluate it on representative image and video diffusion workloads. Compared with fixed-pipeline execution with static parallelism, GF-DiT improves throughput by up to 6.01$\times$, reduces mean latency by up to 95%, lowers SLO violation rates by up to 90%, and reduces communication-group setup overhead from 778 ms to approximately 60 $μ$s.

Grokking is the delayed onset of generalization in neural networks, arising long after they fit the training data. Whether the weight norm causes this delay is disputed: some studies report a critical norm at the transition, others observe grokking with no fixed norm at all. We settle this by intervening on the norm during training rather than only observing it. Under free training with weight decay, networks grok when the weight norm reaches a value Wc that varies little across seeds and learning rates (CV 1 to 2 percent) and grows with the modular base as a power law. When we instead clamp the norm to a fixed multiple rho of Wc and hold it there, the network still groks, but the delay follows T_grok proportional to exp(alpha rho). One exponent, alpha near 7.5, fits this delay across four moduli (R^2 = 0.996). Over the swept ranges the held norm moves the delay by about 19x and the learning rate by only about 2x, and holding the norm above Wc slows grokking rather than preventing it. A final LayerNorm removes the dependence by decoupling weight scale from the network function; without it the exponential law returns. This pinned-norm delay is the exponential counterpart to the logarithmic delay predicted for a freely contracting norm.

This paper presents the participation of team PSL in the QIAS 2026 Shared Task on Arabic Islamic inheritance reasoning. The task evaluates the ability of large language models to solve inheritance cases that require legal interpretation, multi-step reasoning, and precise numerical computation. We compare \textit{commercial} and \textit{open-source} models under a unified prompting strategy to assess their effectiveness in structured legal reasoning with minimal task-specific adaptation. \\ Our results show a clear gap in reliability between the two model families. Commercial models demonstrate stronger performance in identifying eligible heirs, applying exclusion rules, and maintaining consistency across reasoning steps. In contrast, open-source models exhibit greater instability, particularly in cases involving dependent legal decisions and fractional share adjustments. The best performance is achieved by \textit{Gemini 2.5 Flash}, with an MRE of $0.989$.

Automatic speech recognition (ASR) correction has traditionally focused on isolated utterances or short local contexts. However, as text and speech become increasingly interleaved in long interactions, ASR correction requires conversation-level contextual evidence. Existing ASR correction methods often rely on the current hypothesis or concatenate raw dialogue history. In such contexts, sparse correction evidence can be difficult to locate amid redundancy and noise. Addressing these challenges, we propose an ontology memory-augmented ASR correction framework for long text-speech interleaved conversations. The framework organizes preceding interaction history into a dynamically updatable ontology memory, where entities, terminology, surface variants, potential ASR confusions, and semantic relations are stored as retrievable nodes for context-grounded correction. To evaluate this setting, we construct RAMC-Corr, a dataset derived from MAGIC-RAMC for long-range ASR correction with grounded context. Experiments on RAMC-Corr show that our method improves over direct correction in 9 out of 10 paired backbone-setting combinations and encourages more selective and evidence-grounded corrections for context-dependent ASR errors.

Agent-Based Modelling (ABM) relies on clear documentation to ensure credibility and transparency. Although standards exist for documenting models (e.g. ODD), processes (e.g. TRACE, EABSS), and data use (e.g. RAT-RS), their adoption remains limited due to the effort required to produce documentation that is often treated as supplementary. This paper explores the use of Large Language Models (LLMs) to facilitate and partially automate such processes. We conduct a feasibility study focusing on the underused Rigour and Transparency Reporting Standard (RAT-RS), using four LLMs to extract reports from a published ABM paper. We assess consistency and performance across question types, finding that LLMs generate coherent outputs and perform more reliably on descriptive than on explanatory or evaluative tasks. While LLMs can improve reporting quality and consistency, they also exhibit notable limitations. We identify practical heuristics for when LLM-assisted documentation is reliable and when human oversight is needed and call for systematic community-level exploration to enhance rigour and adoption in ABM reporting.

Personalized federated learning (PFL) is one of the important approaches in federated learning for addressing statistical heterogeneity while enabling client-specific adaptation. Many PFL methods split the model into shared and personalized parameters, which are jointly trained on each client. However, this creates an optimization issue: shared parameters are updated by clients optimizing different local objectives, which can lead to inconsistent shared updates and weaken the shared representation. To address this problem, we propose Federated Shared Parameter Correction (FedSPC), a modular correction method for PFL. FedSPC applies control-variate correction only to the shared parameters of a given PFL method, while leaving personalized parameters unchanged. It can be integrated into three common PFL settings: shared feature extractors, shared classifiers, and fully shared models with local regularization. Experiments on CIFAR-100 and Tiny-ImageNet with ViT, ResNet-34, and VGG-11 show that FedSPC improves performance across representative PFL methods, including FedPer, FedRep, FedBABU, LG-FedAvg, and Ditto.

Accurate power estimation is important for understanding and optimizing CPU power behavior, yet practical workflows often rely on simulation-derived information or post-silicon analysis. In this work, we present BigPower, a hierarchical source-level surrogate model for fine-grained module-level power estimation during CPU design. BigPower leverages large language model-based representations together with architectural hierarchy, module connectivity, configuration parameters, and workload context to estimate module-level power consumption directly from source-level design information, without requiring additional simulation during inference. Experimental results in the open-source XiangShan processor family demonstrate practical fine-grained power estimation across diverse configurations and workloads, offering an efficient alternative to conventional simulation-based workflows.

Ultra-long-context capability is becoming indispensable for frontier LLMs: agentic workflows, repository-scale code reasoning, and persistent memory all require the model to jointly attend over hundreds of thousands to millions of tokens, yet the quadratic cost of softmax attention makes this untenable at deployment scale. We introduce MiniMax Sparse Attention (MSA), a blockwise sparse attention built upon Grouped Query Attention (GQA). A lightweight Index Branch scores key-value blocks and independently selects a Top-k subset for each GQA group, enabling group-specific sparse retrieval while maintaining efficient block-level execution; the Main Branch then performs exact block-sparse attention over only the selected blocks. Designed around a principle of simplicity and scalability, MSA is deliberately streamlined, making it straightforward to deploy efficiently across a broad range of GPUs. To translate sparsity into practical speedups, we co-design MSA with a GPU execution path that uses exp-free Top-k selection and KV-outer sparse attention to improve tensor-core utilization under block-granular access. On a 109B-parameter model with native multimodal training, MSA performs on par with GQA while reducing per-token attention compute by 28.4x at 1M context. Paired with our co-designed kernel, MSA achieves 14.2x prefill and 7.6x decoding wall-clock speedups on H800. Our inference kernel is available at: https://github.com/MiniMax-AI/MSA. A production-grade natively multimodal model powered by MSA has been publicly released at: https://huggingface.co/MiniMaxAI/MiniMax-M3.

The ability to resolve complex physical phenomena with high fidelity and at low computational cost is central to addressing key challenges in modern engineering. A prime example lies in hypersonic flows, where the precise prediction of the full flowfield topology, in particular with respect to shock wave location and intensity, is critical. Yet supersonic and hypersonic flows continue to be a stumbling block for traditional reduced-order models and neural emulators that struggle to capture steep gradients in flow states with physical consistency in applications of industrial relevance. To that end, we introduce a fully GPU based workflow that integrates accelerated data generation with the training of neural emulators augmented by uncertainty quantification and physics-aware refinement. Our workflow is enabled by a differentiable high-fidelity solver (JAX-Fluids) which we employ for rapid dataset creation and residual-based improvement of the neural emulator to enhance physical consistency. Building on this framework, we first present a suite of model architectures and analyze their scaling behavior to expose their strengths and shortcomings. We then show that residual-based refinement enables training on cases where only mesh and input parameters are available, substantially reducing residuals and improving physical consistency. Together, differentiable simulation and residual-based refinement yield physics emulators that remain reliable beyond their training distribution, a key requirement for deploying surrogates in real-world engineering design loops.

This paper presents the design, development, and implementation of a specialized forecast-then-optimize algorithmic pricing tool for sales campaigns in fashion e-commerce. Sales events present unique challenges for pricing including volatile demand patterns, rapid pricing decisions, and the need to balance short-term revenue with long-term profitability. We describe our approach combining daily-resolution demand forecasting using gradient-boosted trees with a multi-objective optimization framework that maximizes both long-term profit and net merchandise value for more than 5 million articles. Our solution addresses key limitations of existing weekly-granularity systems by implementing a forecast-then-optimize architecture that reduces pricing decision time from hours to minutes. We validate our approach through 23 A/B tests across 12 markets during 2023-2024 sales campaigns at Zalando, one of Europe's leading online fashion retailers. Experimental results demonstrate that the new pricing system achieves approximately 6% higher profit while maintaining equivalent performance on sales and revenue compared to the previous manual-algorithmic hybrid approach. Based on these results, the algorithm was successfully deployed to production and now handles the majority of algorithmic pricing decisions for sales campaigns at the company.

We introduce the Trajectory-based Quantization Sensitivity Score (TQS), a metric that reframes post-training quantization (PTQ) through the lens of dynamical-systems stability. By modeling the network's rollout as a discrete-time dynamical system, TQS characterizes how quantization-induced errors propagate and amplify over the rollout horizon. Unlike conventional PTQ methods, where sensitivity analysis is often coupled to the quantization procedure, TQS enables a priori sensitivity estimation decoupled from quantizer selection and bit-width assignment. This separation allows for quantization budget planning even for black-box or compiled networks with fused operators. Building on this, we present TQS-PTQ, a flexible mixed-precision framework that requires no calibration data or costly second-order approximations. Our experiments show that a dynamical-systems perspective provides a robust, high-performing pathway for low-precision deployment in resource-constrained settings.

Diffusion large language models (dLLMs) accelerate generation by denoising multiple tokens in parallel, making them attractive for latency-sensitive mobile inference. However, repeated denoising introduces substantial computation on smartphones. Mobile neural processing units (NPUs) offer high-throughput dense matrix computation, but efficiently exploiting them remains challenging: token commitment shrinks per-block effective workloads, token revision complicates KV cache reuse, and limited NPU-visible address space incurs costly remapping and data transfer overheads. In this paper, we propose llada.cpp, the first NPU-aware inference framework for accelerating dLLMs on smartphones. llada.cpp aligns block-wise dLLM inference with the execution characteristics of mobile NPUs through three techniques. (1) Multi-Block Speculative Decoding fills the shrinking workload in late-stage current-block decoding with speculative future-block tokens. (2) Dual-Path Progressive Revision keeps committed tokens revisable until stable and refreshes unstable tokens through a CPU-side path without stalling dense NPU execution. (3) Swap-Optimized Memory Runtime compacts NPU-visible address layouts and overlaps data staging with NPU computation to reduce remapping and transfer overheads. We implement llada.cpp as an end-to-end framework and evaluate it across diverse hardware platforms and dLLM workloads. llada.cpp reduces LLaDA-8B generation latency by 17x-42x over the CPU baseline with prefix KV cache reuse, while preserving generation quality.

This paper examines trade-offs between AI safety and well-being relative to (i) one of the most promising methods for finetuning super-capable AIs, 'Constitutional AI', and (ii) one of the most influential approaches to understanding complex ethical decision making and the conditions for the well-being of rational agents, 'Virtue Ethics'. We finetune various models using a 'Virtuous agent' constitution, a 'Subordinate agent' constitution, and a 'Generic agent' constitution, and evaluate them on 'general safety' (toxic behaviors, misinformation, etc.) and also on their willingness to endorse a wide-range of behaviors that, if adopted by a super-powerful AI, would significantly increase the level of existential risk for humanity. Our results suggest that there is a trade-off between reducing existential risk and reinforcing the beliefs and dispositions that would be conducive to an AI agent's well-being. They also suggest that there is a trade-off between existential risk and general safety: if we finetune an AI to adopt beliefs and dispositions that substantially reduce its existential risk -- by shaping the AI to be systematically subordinate to external human authorities -- we thereby increase the likelihood that a human user can deliberately induce the AI to engage in various kinds of generally unsafe behaviors.

Stream guardrails enable token-level safety detection before full responses are generated. However, they often make overly conservative judgements and block those sensitive but safe tokens, which is known as over-refusal. Due to lack of full context, they also fail to detect implicitly harmful content from jailbreaking. To address these challenges, we propose FreoStream, a novel streaming guardrail framework. Specifically, FreoStream fine-tunes a LoRA module to perform Future-Aware Reasoning when the base guardrail detects unsafe tokens. The reasoning process follows a Future-Reason-Judge paradigm: predict the future, reason about the full context and give the final judgement. This design can effectively reduce over-refusal by incorporating the future information. Moreover, we introduce the Safety-Aligned Optimization module that extracts the safety-aligned component from the reasoning gradients to update the base guardrail model, thereby enhancing streaming safety detection. Extensive experiments on various safety benchmarks demonstrate that FreoStream achieves lower over-refusal rates and better jailbreak defense compared to existing streaming guardrails.

Large Language Models (LLM) have shown impressive capabilities in Register Transfer Level (RTL) code generation, particularly for Verilog. However, evaluating their performance with other Hardware Description Languages (HDL), especially VHDL, remains limited although its distinct language characteristics, such as stricter semantic rules, introduce evaluation considerations that differ from Verilog. This lack of coverage restricts fully understanding of how well current models generalize across hardware design languages with differing structures and semantics. To address this gap, we introduce VHDLSuite, a benchmark-centered infrastructure for scalable VHDL generation evaluation, integrating automated benchmark synthesis, executable validation, and multi-model diagnostic analysis. First, we propose a data pipeline that automatically converts Verilog designs and their accompanying testbenches into executable VHDL benchmark instances, followed by VUnit/GHDL-based validation to ensure each released task is compilable, runnable, and consistently checkable in the VHDL environment. Second, we introduce VHDLBench, a benchmark with over 200 VHDL problems with complete and validated testbenches across a wide range of complexity levels. Third, we extensively evaluate cutting-edge LLMs and uncover key challenges specific on LLM-aided VHDL generation. Our findings provide important insights and support future work in multi-language hardware design automation.Our data pipeline, benchmark, and evaluation framework will be open-sourced.

Recent evidence reported by Tully, Longoni, and Appel (2025) suggests that lower artificial intelligence (AI) literacy predicts greater receptivity toward AI. We revisit this claim using the public data from Study 3 of that article, which measures past usage of five AI tool categories on a five-point frequency scale. We first reproduce the negative association between AI literacy and aggregate AI usage using OLS on participant-level averages, binary logit, ordered logit, and multinomial logit specifications. We then show that the aggregate relationship masks substantial heterogeneity by tool type. In our demographic-adjusted primary specification, AI literacy does not significantly predict text AI usage (ordered-logit $β$ = -0.090, p = .387), whereas it remains a strong predictor of non-text AI adoption ($β$ = -0.377, p < .001). The non-text effect is also robust under Tully et al.'s original Study 3 control specification ($β$ = -0.502, p < .001). Binary, ordered-logit, and multinomial specifications suggest that the non-text relationship is primarily an adoption/non-adoption pattern rather than evidence of intensive use: the demographic-adjusted odds ratio of ever having used a non-text AI tool is 0.68. Thus, in the study that measures self-reported past usage rather than stated preferences, the evidence does not support a simple claim that lower AI literacy predicts greater receptivity to AI in general. It points instead to a narrower pattern of broader adoption across lower-penetration, non-text AI tools.

Multi-agent systems (MAS) were expected to overcome the limitation of single-agent systems (SAS) through collaboration. However, under typicality conditions on the task's constraint graph and bounded inter-agent communication, we prove that the success probability of a MAS is closely tied to the connectivity of task constraints, where each agent has limited information-processing capacity. Specifically, the success probability decays exponentially with an information bottleneck that emerges from partitioning the task's constraint graph among agents. We define this quantity as the \emph{minimum cut cost} $C_{\min}$ of the potential constraint graph of each task. This information-theoretic bound applies to both open systems with external feedback and closed systems without. We validate our theory on both synthetic experiments and real-world empirical data from SWE-bench submissions. From our framework, effective MAS design should incorporate task-inherent constraints alongside engineering optimization, and when $\Cmin$ is high, practitioners should restructure tasks rather than simply scaling agents or communication.

The proliferation of recursive training on synthetic data can alleviate data scarcity but risks model collapse, where repeated training erodes distributional tails and homogenizes outputs. Data selection is widely viewed as a remedy, yet its reliability depends critically on the reference distribution used by the verifier. We show that in low-resource verification regimes, where each verifier observes only a small, fragmented, and biased slice of the target manifold, selection itself becomes biased. This situation naturally arises in low-resource data silos such as healthcare consortia or proprietary financial institutions, where raw data cannot be pooled and local references are inherently incomplete. As a result, selection preferentially retains samples aligned with the local manifold while pruning globally relevant tail modes, turning from a safeguard against collapse into a mechanism that precipitates it. We theoretically prove that such siloed selection accelerates collapse and induces power-law diversity decay. As an initial mitigation, we construct Wasserstein proxy references from multiple silos without sharing raw data. Empirical results confirm that local-reference selection fails on skewed distributions, whereas collaborative proxy references mitigate diversity degradation, suggesting that recursive synthetic-data pipelines require particular caution when real-data coverage is fragmented or scarce.

Business intelligence (BI) increasingly combines dashboard interaction with LLM-based assistance, but these two modes often fall out of sync during multi-step analysis. As users switch between direct dashboard manipulation and natural-language queries, it becomes difficult to preserve a consistent analytical state across filters, hierarchies, metrics, and chart context. We present TwinBI, an agentic digital-twin framework that couples an LLM-based agent system with an executable BI dashboard state. TwinBI unifies conversational interaction, dashboard manipulation, semantic grounding, and provenance tracking through a shared analytical state reconstructed from a unified interaction log. It also exposes artifacts such as schema views, SQL, logs, and an /insights command for state-grounded analytical summaries. We evaluate TwinBI in two complementary ways. In a controlled A/B benchmark with the same backbone agent, TwinBI improves exact-match accuracy from 43.3% to 63.3%, partial-credit accuracy from 48.3% to 70.8%, and substantially reduces timeout rate from 40.0% to 10.0% relative to Dashboard alone. In a usability study, participants benefited from the integrated dashboard-and-chat workflow, with high task accuracy, moderate workload, and favorable ratings for state-aware interaction mechanisms. These results suggest that TwinBI improves both agent-level analytical reliability and user-facing analytical support by turning visible dashboard state into richer actionable context. Our dataset and source code are available at: https://github.com/simonjisu/TwinBI

Evaluating new large language models typically requires costly human annotation campaigns at scale. LLM-as-a-judge offers a cheaper alternative, but judge scores carry systematic errors - such as position bias, self-preference, or intransitivity - that can strongly miscalibrate the resulting rankings. We quantify the resulting judge-human disagreement at two complementary levels. At the local level, we estimate per-battle uncertainty from the judge's own score differences by propagating calibrated win probabilities rather than hard labels into the Bradley-Terry procedure. This alone provides a drastic improvement to Elo estimation accuracy, bringing LLM-derived ratings within 17.9 Elo MAE of human-derived ones when averaged over 55 held-out models on LMArena. At the global level, we apply split conformal prediction to the residual gap between LLM-derived and human-derived Elo ratings across held-out models, producing prediction intervals with distribution-free marginal coverage guarantees that account for irreducible LLM-human disagreement. Together, these two layers yield a low-cost evaluation tool that provides developers with calibrated Elo estimates and honest uncertainty bounds, without access to large-scale human annotations. To facilitate reproducibility, we release our code at https://github.com/kargibora/SoftElo .

Recent AI systems can generate texts, software architectures, hypotheses, designs, and scientific workflows that appear creative. This paper asks under what conditions a machine can be called genuinely creative, and how human agency can be preserved within shared cognitive and creative environments. It develops a requirement framework derived from Designics, the science of meaning-bearing intentional change. The paper argues that genuine machine creativity should not be defined by output novelty, current performance, or transient architecture alone. Instead, creativity is understood as the structural transformation of incomplete situations through recursive intervention dynamics. On this view, it depends on ten requirements: environment representation, scoped perception, conflict identification, intervention capability, consequence observation, knowledge and environment update, rescoping, local-to-global unfolding, value-based scoping, and human-AI co-living. These are organized through the three laws of Designics: perception, conflict, and capability. The paper illustrates the computational tractability of these requirements through selected cyber-physical and cyber-biological studies, including recursive element extraction, autonomous mesh generation, and neurophysiological and workload analysis. It then treats open-ended systems, automated discovery frameworks, self-modifying agents, foundation models, and agentic workflows as pressure cases: they demonstrate powerful generative means but do not by themselves establish genuine machine creativity. Finally, the paper argues that proactive AI ethics is internal to genuine machine creativity rather than an after-the-fact filter. Value-based scoping and human-AI co-living must shape how creative machines perceive environments, identify conflicts, select interventions, observe consequences, update knowledge, and rescope future action.

Spatio-Temporal forecasting is crucial in diverse fields, such as transportation, climate, and energy. Urban spatio-temporal data exhibits temporal mirage: similar short-window inputs have divergent future trends, and vice versa. Existing spatio-temporal graph neural networks (STGNNs) cannot effectively identify such mirages. We argue that the core reason lies in the short-window inputs that have incomplete period observation, heterogeneous global spatial correlation, and cross-period superposition causality. To bridge this gap, we develop a novel Multi- Period Pattern Pre-training (MP3), a plug-and-play pre-training plugin for distinguishing temporal mirages. MP3 presents two core innovations: (1) The multi-period pattern learning is designed to learn multi-period patterns from long time series. Specifically, multi-period temporal modeling leverages edge convolution to identify different multi-period patterns. Multi-period spatial modeling uses a bottleneck project and a global memory bank to capture heterogeneous global spatial relations efficiently. Cross-period pattern interaction employs a causality-enhanced Transformer to capture dependencies across different period patterns. (2) This plugin can seamlessly integrate into existing STGNN backbones to strengthen their forecasting performance. The experiment on five STGNN baselines across five real-world datasets (including a large-scale dataset CA) verify the effectiveness, superior scalability and strong adaptability of MP3, which brings consistent and robust performance improvements across all evaluated baselines. On average, MP3 reduces the MAE 4.7% and the RMSE 5.0%. The code can be available at https://github.com/YAN-outlook/MP3.

Real memory chip characterization yields insights into fundamental operational characteristics of modern memory, enabling new mechanisms that improve memory performance, robustness, security, and energy efficiency. We describe our large-scale DRAM characterization laboratory for understanding DRAM. A key building block of this laboratory is DRAM Bender, a versatile and easy-to-use modern DRAM characterization infrastructure. We have updated DRAM Bender to i) introduce support for new types of characterization experiments, ii) expand on its DRAM interface standard support, and iii) make it easier to use at large scale. This paper introduces these updates for the first time. We hope our infrastructure enables the community to discover new problems and solve critical memory scaling issues, enabling the overcoming of the huge memory bottleneck that plagues modern computing systems.

Large language models (LLMs) exhibit exceptional general language processing capabilities, but their memory and compute costs hinder deployment. Ternarization has emerged as a promising compression technique, offering significant reductions in model size and inference complexity. However, existing methods struggle with heavy-tailed activation distributions and therefore keep activations in high precision, fundamentally limiting end-to-end inference acceleration. To overcome this limitation, we propose TWLA, a post-training quantization (PTQ) framework that achieves 1.58-bit weight compression and 4-bit activation quantization while maintaining high accuracy. TWLA comprises three components: (1) Euclidean-to-Manifold Asymmetric Ternary Quantizer (E2M-ATQ) minimizes layer-output error under weight ternarization via a two-stage optimization from Euclidean initialization to manifold relocation; (2) Kronecker Orthogonal Tri-Modal Shaping (KOTMS) applies a Kronecker-structured orthogonal rotation to reshape weights into ternary-friendly tri-modal distributions, while the shared rotation statistically suppresses activation outliers; and (3) Inter-Layer Aware Activation Mixed Precision (ILA-AMP) explicitly introduces adjacent-layer second-order interaction costs in bit allocation and jointly optimizes for the layer-wise disparity of activation quantization gains induced by the shared orthogonal transform, preventing cascades triggered by a few weak layers. Extensive experiments demonstrate that TWLA maintains high accuracy under W1.58A4, while delivering significant inference acceleration. The code is available at https://github.com/Kishon-zzx/TWLA.

Intersection-over-Union (IoU), as a pivotal metric for evaluating the spatial alignment between candidate proposals and ground-truth annotations, directly determines the quality of positive sample sets and the training efficacy of visual detection models. Through theoretical modeling and analysis, we uncover a non-sensitive region on the IoU response curve, within which samples yield nearly identical IoU scores despite distinct geometric overlaps. To overcome this limitation, we introduce a set of morphological similarity metrics covering area, shape, and aspect ratio, to refine the positive sample assignment process, thereby ensuring more discriminative and reliable matching. A supplementary matching score is derived via mean-based aggregation of these multidimensional similarities, compensating for the intrinsic limitation of IoU in representing structural correspondence. Theoretically, incorporating morphological similarity reshapes the response distribution of the matching function, yielding both effective directional gradients and polygon-like iso-response contours, which tightly confine high-response regions around each ground-truth instance and substantially enhance the precision of positive sample selection. Experiments based on the YOLOv9 framework demonstrate consistent performance gains on both NEUDET and GC10- DET datasets. Notably, the proposed approach is fully plug-and-play and incurs zero additional inference overhead, thereby ensuring deployment efficiency for industrial visual inspection.

This paper introduces YeasierAgent, an application-building paradigm based on symbiotic agents, narrative worlds, and scene-aware interaction. It challenges the conventional device-coupled model of software by redefining applications as collaborative spaces among users, agents, and worlds. We present a system architecture that achieves two primary contributions: (1) enabling the rapid, cross-platform construction of agent-native applications by utilizing platform-agnostic interactive units (agents, scenes, dialogue) rather than fixed graphical layouts; and (2) unifying the emotional companionship and practical tool execution attributes of intelligent agents within a single experiential sandbox. By integrating automated generation, user-created worlds, and spatial multi-agent collaboration, YeasierAgent formalizes the category of Symbiotic Agent-Native Applications, demonstrating a shift from isolated, tool-specific chatbots toward cohesive, socially embedded computational environments.

Data-driven engineering design is constrained by the lack of large-scale 3D datasets that pair geometry with physics-based performance labels. In particular, existing 3D data augmentation techniques have limitations in preserving subtle and diverse geometric variations, and it remains difficult to automate the subsequent simulation-labeling process, where boundary conditions vary depending on the generated geometry. We present DeepJEB++, a foundation-model-driven data-augmentation framework that expands a small seed set of jet engine brackets into a large, simulation-labeled 3D dataset under constrained resources. Our key idea is to augment in the data-rich 2D latent space, then transfer to 3D. In Stage 1, we fine-tune a pretrained 2D latent diffusion model on multi-view renders and synthesize novel views by latent interpolation, retaining manufacturable designs through a vision-language-model (VLM) quality filter. In Stage 2, the validated images are lifted to 3D meshes by a domain-adapted generative foundation model. In Stage 3, an automated pipeline recognizes the load and bolt interfaces on each mesh and assigns finite-element labels -- mass, stress, and displacement -- without manual intervention. We assess augmentation quality along three intrinsic axes: manufacturability, label fidelity against the SimJEB ground truth, and distributional consistency. Starting from fewer than 400 seed designs, DeepJEB++ yields 15,360 simulation-labeled 3D brackets -- a 40x expansion -- using a single GPU per stage. The dataset will be made publicly available to support reproducible engineering-AI research.

Arditi et al. (2024) has shown that refusal in safety fine-tuned chat models is mediated by a single linear direction in the residual stream, recoverable by a difference-in-means (DiM) of harmful and harmless activations. We compare DiM-based interventions (activation addition and directional ablation) with two interventions derived from Iterative Nullspace Projection (INLP) -- nullspace projection and counterfactual flipping -- on five open-weight chat models, asking whether INLP can match DiM at steering refusal and whether its richer parameterisation yields more tweakable interventions. INLP counterfactual flipping is competitive with DiM directional ablation on refusal suppression, while nullspace projection is consistently weaker. Restricting INLP to the leading directions of the extracted subspace preserves most of the suppression effect at near-baseline perplexity, giving a tunable capability. Geometrically, the two INLP interventions land in qualitatively different regions of activation space: nullspace projection collapses transformed activations \emph{between} the harmful and harmless clusters, while counterfactual flipping moves them into the opposite cluster, suggesting that the model encodes the absence of a concept differently from its opposite -- an intriguing distinction that warrants further investigation in future work.

When a user reveals task-critical information across several conversation turns, LLM accuracy drops by up to 65% despite full context availability. We show that this Lost in Conversation degradation can be substantially mitigated by training models to maintain a compact rolling memory instead of attending to a growing history. To make such training scalable, we introduce a low-cost sharding pipeline that converts single-turn QA datasets into multi-turn fragmented-information episodes, eliminating the need for hours of manual annotation. Training only on sharded GSM8K, our memory-augmented policy significantly improves multi-turn accuracy and generalises zero-shot to harder math and out-of-domain long-context QA. Moreover, memory-trained models outperform full-history baselines even when given the full history at test time, suggesting that learning to compress induces more robust incremental reasoning than full-context exposure alone.

AI alignment, interpretability, steering, and neural perturbation studies identify order-inducing objects. We argue that order is not control. Control requires a receiver-gated response law: a denominator-indexed operator mapping material state, action/drive, bath, and receiver state to response displacement, sinks, effort, and basin projection. We identify it across biological, LLM, adapter, and stochastic-operator panels. The laws are local: an intervention can be admitted, saturated, sign-changing, leaky, or overdriven depending on medium, bath, receiver state, action port, and comparator. Control is assigned when finite effort moves a target or outcome-readout class under the same denominator while damage, null/evasive, invalid format, overdrive, and unnecessary effort stay bounded. Mouse ALM, C. elegans, and zebrafish panels provide physical response-operator evidence while excluding coordinate identity and controller conclusions. LLM panels show generated-output response laws: across four material conditions, response vectors are predictable at 72.8-73.7% component-sign accuracy, rising to 84.3-84.8% on nonzero components; held-out observers predict system-effect and target/oracle families at 93.6% and 91.7% accuracy. Constitution-conditioned adapters reshape susceptibility as prepared media, and stochastic-operator panels separate measured opportunity from deployable action policies. This gives a driven-dissipative response-system account at the mesoscopic control level: drives act through prepared media, baths, and receivers, producing admitted movement, impedance, sinks, or overdrive. The evidence supports local admitted control and measurable stochastic response operators, while leaving deployable pre-generation control, hidden/logit causal sufficiency, biological-to-LLM coordinate identity, and literal thermodynamic quantities outside scope.

Hierarchical multi-agent systems (MAS) are rapidly being deployed in high-stakes workflows across domains such as finance and software engineering. In these systems, safety and security are inherently distributed across role-specialized agents, significantly expanding the attack surface, particularly under coordinated adversarial behaviors such as privilege escalation and cross-agent collusion. Existing red-teaming approaches for MAS remain limited: they rely on heuristic selection of target agents and perturb isolated message streams, leaving critical questions unanswered as which agents are most responsible for system safety, and how compromised agents can coordinate to bypass defenses. We propose MAStrike, a closed-loop framework for collusive red-teaming in hierarchical MAS. We propose the first agent-level Shapley value analysis for MAS, quantifying each agent's marginal contribution to system robustness under task-specific distributions. GGuided by this attribution, MAStrike identifies vulnerable agent coalitions and generates coordinated, role-aware adversarial manipulations. These attacks are iteratively refined through structured causal diagnosis, attributing failure cases to uncompromised agents that block adversarial attempts. We further build a comprehensive MAS red-teaming benchmark and controllable environments spanning diverse hierarchical topologies and domains, including finance, software engineering, and CRM. Extensive experiments across MAS built on multiple frontier models show that MAStrike substantially outperforms heuristic baselines. Our analysis further uncovers non-trivial Shapley value distributions and higher-order interaction structures among agents, revealing critical vulnerabilities and coordination patterns that are overlooked by prior single-agent or template-based methods.

For robotics to be effectively integrated into household or industrial environments, machines must adapt to natural-language prompts in real time. Although Vision-Language Models (VLMs) have enabled zero-shot generalization in robot task and motion planning (TAMP), current state-of-the-art approaches often remain computationally "heavyweight" or require extensive training on thousands of demonstrations. We present GRASP (Grounded Reasoning and Symbolic Planning), a framework designed as a step toward open-vocabulary tabletop manipulation. Our approach leverages a pretrained VLM to translate natural-language queries into neuro-symbolic goal states, grounded in the physical world via a bounding-box detection pipeline. Unlike methods that rely on fixed color lists or hard-coded coordinates, GRASP enables robots to interpret abstract spatial concepts such as "top shelf" and execute tasks without additional fine-tuning. We achieve 73.3% overall success across 90 real-robot trials at three difficulty levels, requiring no task-specific training.

We present an approach to fine-tuning large language models using Direct Preference Optimization (DPO), a reinforcement learning technique. Our experimental results demonstrate that DPO simplifies the training pipeline, improves computational efficiency, and achieves competitive performance. The evaluation using BLEU, ROUGE, and cosine similarity metrics indicates effective learning and convergence, though further investigation is needed to address observed training instability.

Understanding the digital world on mobile devices is shifting from static UI perception to dynamic action comprehension. This capability enables models to convert visual state transitions into operational knowledge, defined as short natural-language sentences that describe action types, target UI elements, textual arguments, and execution orders. However, due to the highly diverse and heterogeneous UI designs across applications, existing vision-language models (VLMs) struggle to accurately infer these underlying operations. To bridge this gap, we introduce Teach VLM, a core model designed to translate mobile screen trajectories into step-wise operational knowledge by extracting and analyzing operation-related keyframes from demonstration videos. To address the scarcity of aligned training data, we develop a systematic data flywheel for scalable data acquisition. We further introduce a novel Chinese Mobile Screen Teach Benchmark for fine-grained evaluation. Building upon Teach VLM, we propose the Teach-and-Repeat paradigm, where the generated operational knowledge serves as an interpretable procedural reference to guide downstream screen-based execution agents. Extensive evaluations demonstrate that Teach VLM significantly outperforms strong VLM baselines, achieving state-of-the-art performance in operation semantics prediction. Furthermore, experiments in Android World show that our paradigm yields consistent Task Success Rate improvements for downstream agents. Together, Teach VLM and the Teach-and-Repeat paradigm offer a practical pathway from raw demonstrations to reusable task automation.

Privacy auditing provides an important safeguard by estimating the actual information leaked by a model, thus ensuring that theoretical privacy guarantees hold in practice. We study empirical privacy auditing for differentially private (DP) machine learning, focusing on efficient one-run methods for mechanisms such as DP-SGD. Prior one-run approaches threshold training examples or "canaries" into binary membership guesses, which discards useful information. We show that, in the white-box DP-SGD setting, canary-aligned signals naturally form a sequence of random variables whose normalized sum is asymptotically Gaussian. Leveraging this distributional perspective, we develop a DP-auditing framework that leads to tighter privacy lower bounds from a single training run.

Most learned dexterous grasp generators relegate contact forces to a downstream verification step, so a kinematically-plausible pose can still violate the conditions for a stable physical grasp. We address this with EquiDexFlow, an SE(3)-equivariant flow-matching model that jointly predicts wrist pose, joint angles, fingertip contacts, surface normals, and contact forces from an object point cloud. Our architecture projects contacts onto the object surface and forces into the Coulomb friction cone by construction, so placement and friction compliance hold without loss penalties. We prove end-to-end SE(3) equivariance and verify it empirically over 200 rotations, with wrist residuals below $0.04^\circ$ and exactly zero joint deviation. Trained on 8,100 force-closure grasps across 81 objects for the 16-DoF Allegro Hand, our model achieves zero friction violations, the best composite score, and the lowest wrench residual among all ablation variants. We retarget decoded fingertip contacts to a 16-DoF LEAP Hand via per-finger inverse kinematics, and our hardware-feasible refinement places every joint at least 5% inside its actuator envelope while preserving wrench balance. On the physical robot, retargeted EquiDexFlow-decoded grasps complete open-loop pick-and-hold trials on all six test objects, with every asymmetric object succeeding at both the canonical pose and a $120^\circ$ co-rotation. Videos, code, and checkpoints are available at https://equidexflow.github.io.

The best agent on WorkBench in March 2024, GPT-4, completed 43% of tasks and took an unintended harmful action, such as emailing the wrong person, on 26% of them. We re-visit the benchmark in June 2026 and find that the best agent to date, Claude Opus 4.8, completes 89% and takes an unintended harmful action on 2.5%. Aside from this considerable progress in frontier agent performance, three things stand out. First, capability and safety go together on WorkBench rather than trade off, so the models that finish the most tasks also do the least unintended damage. Second, while several classes of error have been totally eliminated, frontier models still make some basic mistakes that occasionally result in irreversible harm, such as sending an email to the wrong person. Third, the rise of open-weight models has drastically lowered costs for a performance level that was previously only accessible to proprietary models, while frontier costs have stayed relatively stable. We release an updated version of the benchmark with data and code quality improvements, new model scores, and analysis of agent progress on WorkBench since 2024.

Predicting cellular transcriptional responses to genetic perturbations is a central problem in single-cell biology, especially in the zero-shot setting where the perturbed gene or gene combination is unseen during training. A major difficulty is that perturbation effects are not determined by expression state alone: they depend on how the perturbed gene product influences other genes and proteins, how those downstream factors act on cis-regulatory elements, and which regulatory programs are active in the current cell state. To better capture this biological complexity, we propose CisTransCell, a cell-conditioned multi-modal framework for single-cell perturbation prediction that augments each gene with two complementary priors: a regulatory-sequence prior that captures how the gene is controlled, and a coding-sequence prior that captures what the gene product does. By integrating these priors with cellular expression state, CisTransCell models perturbation response as a cascade from gene function to regulatory control to downstream transcriptional change. Experiments on benchmark single-cell perturbation datasets show that CisTransCell achieves strong performance in zero-shot perturbation prediction.

Automated analysis of K-12 classroom dynamics faces challenges due to background noise and variable child speech, often confounding acoustic-only models. This study evaluates a multimodal speaker identification framework anchoring acoustic embeddings with LLM-derived semantic context. Using a subset of the EDSI dataset (8 math classrooms, N = 2,801 utterances), we found an acoustic baseline (ECAPA-TDNN) achieved only 39.0% accuracy. By integrating transcript-based "contextual anchoring" into a gradient boosting classifier, our multimodal approach raised student identification to 50.3%. Performance also improved for utterances over 5 seconds, reaching 76.9% accuracy (vs. 64.9% baseline) with a 90.9% Top-3 accuracy. Additionally, the model distinguished teacher vs. student roles with 99.3% accuracy. This approach advances the feasibility of automated feedback systems capable of considering individual student participation, a crucial step for supporting equitable instruction at scale.

Language-model post-training is the main stage at which model behavior is shaped, yet it still largely involves optimization of scalar rewards that summarize diverse desiderata. This abstraction gives practitioners little visibility into what their data actually teaches models, allowing spurious correlations to be learned by a model and inducing undesirable behaviors such as over-stylization and sycophancy. To address this problem, we ask: can we inspect a preference dataset before optimization and decide, at the level of concepts, which behaviors a model should be allowed to learn? Motivated by this, we introduce a data-centric post-training pipeline that uses interpretability protocols to develop statistical hypotheses for the latent concepts separating preferred from dispreferred generations, making them explicit for fine-grained user feedback. Building on this view, we unify several interpretability-based training protocols as ways of shaping rewards via feature or data interventions. Empirically, we show that our pipeline diagnoses undesirable signals in existing preference data, mitigates off-target learning, and can also help amplify or shape desired properties such as safeguards and model personality. More broadly, our results suggest that interpretability can turn post-training from optimizing opaque proxy rewards into a process of auditing and sculpting the learning signal itself.

We study post-training quantization (PTQ) of Ideogram 4.0, a 9.3B flow-matching diffusion transformer (DiT) that realizes classifier-free guidance with two separate-weight copies of a single-stream backbone and is conditioned by a Qwen3-VL text encoder, targeting Ampere RTX~3090 GPUs, which lack FP8 tensor cores. Because Ideogram~4.0 is trained on structured JSON captions, we evaluate every variant under schema-valid JSON prompts produced by an LLM expander built to Ideogram's published caption specification, and score them with a battery spanning human-preference (HPSv2), CLIP, and PickScore for standalone quality; PP-OCR exact-match and edit distance for text; and PSNR/SSIM/LPIPS for fidelity to the FP8 reference (the highest-precision public checkpoint) output. On a 300-prompt benchmark with paired bootstrap confidence intervals, an INT8 W8A8 recipe (per-channel weights, per-token dynamic activations, SmoothQuant, and bf16 protection of a small high-fragility layer set) is statistically indistinguishable from FP8 on CLIP and PickScore (paired CIs include zero) and within ~0.004 HPSv2, and, at its 8-bit size, is the most faithful reproduction of the FP8 output (LPIPS 0.243 vs 0.277/0.306 for the half-size 4-bit baselines; the INT8-Q4_K gap excludes zero). A GGUF Q4_K quantization reaches the same standalone quality as the published NF4 baseline at the same on-disk size, making it the Pareto choice on the quality-memory frontier. We further show that under JSON prompts all four variants reach parity on standalone quality, the variants separate on fidelity and text rendering, not on aggregate image-quality scores, and that text legibility, near-zero when the model is prompted with raw strings, reaches 55% OCR exact-match under the JSON captions it expects. We release the INT8 W8A8 and GGUF Q4_K quantized weights on Hugging Face under a gated, non-commercial license.

Deep research and agent evolution serve as de-facto tasks for AI agents in real-world applications toward artificial general intelligence. The former enables autonomous retrieval and integration of information in open-ended environments to tackle open-ended research tasks, yet it is constrained by the static parametric deep research capabilities of agent systems. The latter allows agents to autonomously interact with the environment to gain experiences that evolve model capabilities. However, its effectiveness has been widely validated only on verifiable tasks with standard answers, leaving a gap with open-ended research tasks. To bridge these two critical tasks, we propose the Hybrid Open-Ended Tri-Evolution (HOTE) framework, which leverages hybrid-mode reinforcement learning to facilitate the collaborative evolution of a proposer, solver and judge based on web-scale knowledge, moving toward autonomous evolving agents in open-ended tasks and environments. Extensive experiments on three long-form deep research benchmarks demonstrate that the 8B model trained via HOTE surpasses the strongest static open 8-32B models as well as those trained by state-of-the-art deep research training methods with less time overhead, and further verify that the evolution of all three modules in HOTE is indispensable.

This paper presents a novel data-free and training-free compression approach for speech foundation models using channelwise clustering via k-means. More fine-grained, mixed sparsity pruning by layer-level varying number of parameter clusters is also explored. Experiments conducted on the LibriSpeech dataset suggest that when operating with pruning sparsity of 50% on HuBERT-large, consistent WER reductions of 27.73%/18.61% absolute (34.37%/21.91% relative) over the magnitude-based pruning were obtained on the test-clean and test-other subsets before fine-tuning and 0.19%/0.79% absolute (3.36%/4.62% relative) after fine-tuning with only 3 epochs. Similar WER reductions of 2.86%/5.02% absolute (59.21%/55.29% relative) were observed against magnitudebased pruning on Whisper-large-v3 at 10% sparsity, all with no significant WER increase relative to the uncompressed baseline.

We study controlled post-training refusal suppression in routed MoE and hybrid-MoE foundation models, aiming to increase non-refusal target-response behavior while preserving general capability under a compact intervention footprint. Existing broad direction-based edits can perturb general-purpose computation, whereas support-only expert edits often lack sufficient capacity to correct heterogeneous refusal representations. To address this limitation, we introduce Localized Multidirectional Correction (LoMC), a support-gated intervention framework that follows a support-then-correction execution order: it first identifies a compact edit support, then aggregates prototype correction directions into layer-wise correction directions, and finally applies rank-one layer-wise correction only within the selected support. By using the edit support as a structural gating constraint, LoMC increases correction capacity without expanding the intervention scope. Experiments on text-only and multimodal safety benchmarks across four routed backbones show that LoMC substantially improves non-refusal target-response behavior while maintaining general capability under a compact intervention footprint.

Multi-chiplet GPUs scale compute throughput and high-bandwidth memory (HBM) capacity, but their non-uniform memory system makes locality between chiplets and their data critical to the GPU's performance and energy efficiency. Locality-aware scheduling and data placement identify which data should reside near each chiplet. However, in general matrix multiplication (GEMM), locality-aware data placement often becomes incompatible with a fixed page-granularity data interleaving, since the optimal granularity for mapping data across chiplets varies widely across workloads. We propose Chiplet-Contiguous Layout, a global memory layout that stores chiplet-local data contiguously. Chiplet-Contiguous Layout enables locality-aware placement compatible with page-granularity placement across diverse large language model (LLM) GEMM shapes, without changes to the operating system or hardware. On representative LLM inference and training GEMMs from Qwen 3 30B and Llama 3.1 70B, Chiplet-Contiguous Layout on average reduces remote HBM traffic by 13.0x on Qwen and 20.7x on Llama over 4\,KB interleaving, and by 3.3x and 3.7x over coarse locality-aware placement.

Multi-chiplet GPUs split memory into local and remote HBM regions across a silicon interposer, and reducing the remote HBM traffic is crucial for the performance and energy efficiency of multi-chiplet GPUs. For general matrix multiplication (GEMM), the dominant operator in large language models (LLMs), the resulting inter-chiplet traffic depends strongly on kernel choices such as operand layout, CTA traversal order, and data placement, and the optimal strategy to minimize remote accesses is nontrivial. We present a fast, functional, tile-level locality simulator that models CTA scheduling, per-chiplet L2 caches, and local/remote HBM accesses to evaluate a full-size LLM GEMM configuration. Across representative LLM GEMMs, the simulator shows that remote traffic varies by up to 58x across the design space for the same GEMM dimensions. Moreover, using the simulator as feedback, an agentic AI discovers that a 2D block-swizzle CTA traversal reduces remote traffic over the best 1D traversal by up to 5.1x under round-robin placement, identifying CTA traversal order as a first-order, GEMM-dependent design knob for inter-chiplet traffic.

Token-level hallucination detectors are evaluated as classifiers, by AUC over all tokens, yet a streaming monitor is judged by its reaction time: the number of tokens that pass between the onset of a hallucination and the alarm. We formulate hallucination onset detection as a quickest change detection problem. A first-order Markov model of the latent faithful/hallucinated state, validated on RAGTruth, places the task inside classical change-point theory and yields Lorden's lower bound on detection delay: about 1.3 tokens at a false-alarm rate of 0.01. We then show that a causal recurrent labeler acts as a CUSUM with a learned increment. Among the onsets it catches it detects in 11-13 tokens, against 31 for a linear per-token baseline, though at this false-alarm budget every detector catches under a third of onsets and the recall-honest delay is 56-66 tokens: low-false-alarm onset detection is hard. A controlled decomposition attributes the speed advantage mostly to a better per-token score rather than to temporal accumulation. An information-rate optimality theorem of Donsker-Varadhan type explains the remaining order-of-magnitude gap: the learned score realizes only 1/4.5 of the divergence the features carry, a deficit that recalibration cannot remove, with the remainder a finite-horizon effect. Classification metrics conceal this delay structure; sequential analysis makes it measurable.

RDMA one-sided verbs are the natural primitive for memory disaggregation, but they require the client to supply the exact remote address. The 1-RTT performance breaks down when the target address depends on data that must first be read from remote memory, a pattern we call the Indirection Wall. Indirection is pervasive: graph traversals follow pointers hop by hop, address translation walks multi-level page tables, distributed coordination requires conditional multi-host logic, and disaggregated LLM inference must resolve paged KV caches through block-table lookups. Each level of indirection costs one sequentially dependent network round-trip, yet offloading to existing RDMA NICs either consumes remote CPU cycles or has limited throughput. We present Tiara, a compact, statically verifiable instruction set that executes on the memory-side NIC. Tiara operators are pre-registered programs, analogous to eBPF programs in the kernel, that resolve indirection locally, collapsing multi-RTT dependent chains into a single round-trip. On an FPGA-based prototype, Tiara reduces 10-hop graph-traversal latency by 2.85x over one-sided RDMA while sustaining 3.4x higher throughput, cuts page-table walk latency by 62%, reduces uncontended distributed-lock latency by 2.9x, achieves 2.8x throughput for disaggregated PagedAttention at 8 KB blocks, and 1.88x MoE expert-gather latency at 32 experts.

The recent success of agent swarms has shifted the paradigm of large language model (LLM)-based agents from single-agent workflows to multi-agent systems, highlighting the importance of agent orchestration for task decomposition and collaboration. However, existing orchestration frameworks are limited to a narrow set of modalities and struggle to generalize to more complex settings where heterogeneous modalities coexist and interact. This limitation becomes particularly pronounced in omnimodal scenarios, where tasks require the unified understanding and coordination of diverse inputs such as text, image, audio, and video. In this work, we propose Orchestra-o1, an omnimodal agent orchestration framework designed to support efficient agent collaboration across multiple modalities. Orchestra-o1 introduces a unified orchestration mechanism that enables modality-aware task decomposition, online sub-agent specialization, and parallel sub-task execution. This scalable design allows agent systems to effectively tackle complex real-world tasks involving heterogeneous information sources, surpassing the second-best approach by 10.3% accuracy on the OmniGAIA benchmark. Furthermore, we introduce decision-aligned group relative policy optimization (DA-GRPO), an efficient agentic reinforcement learning approach for training Orchestra-o1-8B, which also achieves state-of-the-art performance against all existing open-source omnimodal agents.

Focused helium ion beam (He-FIB) irradiation with 30 keV ions is a key tool for nanoscale patterning and defect engineering in high transition temperature (Tc) cuprate superconducting devices, yet its usable fluence window is constrained by the competing requirements of reliable superconductivity suppression and minimal structural degradation. In this work, we provide a comprehensive dataset on the effects of large-area 30 keV He+ ion exposure on the electric transport and superconducting properties of epitaxial YBa2Cu3O7 (YBCO) thin films. X-ray diffraction shows a fluence-driven loss of crystalline order accompanied by an out-of-plane lattice expansion and an orthorhombic-to-tetragonal transition, culminating at predominant amorphization at the highest fluence of 1 x 10^16 cm^-2. Raman spectra exhibit increasing disorder while lacking signatures of oxygen depletion, indicating that irradiation mainly generates oxygen-related Frenkel defects rather than changing the carrier concentration. Consistently, with increasing fluence, the normal-state resistivity ρ_N(T) at temperature T above Tc increases strongly, while d ρ_N/d T remains nearly unchanged at moderate fluence. The suppression of Tc is accurately described by Abrikosov-Gor'kov pair breaking and reaches complete quenching of superconductivity at 4.5 x 10^15 cm^-2. The anisotropic upper critical fields decrease approximately exponentially with increasing fluence, the vortex activation energy is reduced, and the anisotropy drops, in contrast to oxygen-depleted YBCO. Hall-angle analysis confirms a nearly constant carrier density but a systematic increase in defect scattering and reduced mobility, consistent with a crossover toward the dirty limit at high fluence. These results establish quantitative fluence thresholds and a practical operational window for He-FIB nanopatterning of YBCO quantum circuits.

The theory of wave turbulence describes the properties of physical systems composed of a set of weak-amplitude random waves interacting nonlinearly. Here, we study odd wave turbulence, which arises in chiral media subjected to non-reciprocal stresses, notably odd viscosity and odd elasticity. In both cases, we consider simple models for which we can derive and solve analytically the kinetic equations describing the long-term statistical behavior of spectral quantities such as energy or wave action. For odd viscosity, we consider a three-dimensional model that exhibits wave turbulence involving three-wave interactions, which gives rise to a direct energy cascade characterized by an anisotropic Kolmogorov-Zakharov (KZ) spectrum. For odd elasticity, we consider a quasi-one-dimensional overdamped model that exhibits much slower dynamics involving six-wave interactions. In that case, the KZ spectrum corresponding to a forward cascade of a conserved quantity we call odd energy, is nonlocal and therefore does not constitute a physical solution. However, the other KZ solution, which describes an inverse cascade of wave action, is only marginally non-local and is therefore valid up to a logarithmic correction. These two analytical theories provide a rigorous interpretation of direct numerical simulations, where the KZ spectrum is observed both in the case of odd viscosity (forward cascade) and of odd elasticity (inverse cascade).

A large body of experimental studies of swift heavy ion latent tracks in dielectric materials has produced a wide range of estimates of track size. We investigate the physical meaning of these estimates by examining the different criteria of track boundary probed by various experimental techniques, including SAXS, XRD, chemical etching and conductometry. We show that different methods probe different physical aspects of ion induced modification, such as electron density redistribution, molecular ordering, chemical reactivity and charge separation, resulting in different determinations of effective track boundaries. Particular attention is paid to polymer films with electret-like properties, where post irradiation redistribution of weakly bound electrons may play an important role in the evolution of latent track structure.

Minimising bacterial surface adhesion and subsequent biofilm formation in industrial and medical settings requires understanding how bacteria are transported and adhere to complex surface geometries in the presence of non-uniform flow. In this paper, we consider the transport of a dilute suspension of motile bacteria through a corrugated two-dimensional channel with perfectly adhesive walls. We asymptotically analyse the diffusive boundary layer that forms in high velocity flows using a curvilinear coordinate system based on the fluid streamfunction, presenting a similarity solution to the diffusivity-varying diffusion-type equation that arises. From this solution, we derive an analytical expression for the bacterial adhesion rate as a function of surface arclength and the spatially varying wall shear rate. Our model predicts that bacterial adhesion becomes localised on curved surfaces, with bacteria showing preferential adhesion to wall `peaks' at lower shear rates and preferential adhesion to wall `valleys' at higher shear rates. More broadly, our results highlight how spatially varying flows generated by complex geometries can lead to localised bacterial adhesion, with potential implications for both enhancing and minimising biofilm formation.

A central problem in statistical physics is to extend the Boltzmann distribution to nonequilibrium steady states (NESS). We prove that NESS on large dense networks have Boltzmann-like occupation despite extensive entropy production. We further show that the active-matter heuristic of "low rattling" is asymptotically exact. Intuitively, these NESS spend a greater fraction of their time in states they leave more slowly. This explanation extends to the broader class of "equiaccessible" steady states, which play a role in our analysis akin to that of equilibrium in linear response.

Light-matter interactions are often modeled via the Dicke model, namely, by two-level systems coupled to a cavity mode. Alas, the threshold for superradiance is often experimentally inaccessible or hindered by light's diamagnetic term. Here, within the Dicke setting, we consider self-interacting light in a cavity, modeled by a photonic Kerr nonlinearity. We show that negative Kerr nonlinearity gives rise to a low-threshold superradiant phase with spin inversion. While unstable in a closed system, cavity dissipation stabilizes this lit phase, opening avenues for lasing and bath-engineered phases.

The covariant formulation of stochastic thermodynamics requires treating the stochastic work as a 4-vector, posing significant challenges for quantum systems due to the non-commutativity. We introduce a new quasiprobability distribution for the work 4-vector, which combines the Wigner and Margenau-Hill quasiprobabilities. This extends the covariant fluctuation theorems from classical to quantum regime. We illustrate our findings with a scalar field driven by classical particles with a generalized version of trace formula. Our work establishes a quasiprobability approach to studying relativistic quantum thermodynamics in a covariant way.

Tuning spin-orbit coupling is essential in controlling both spin and charge in confined semiconductor nanostructures, yet it is rarely a truly controllable parameter. Here, we show control over the spin-orbit Hamiltonian in quantum dots and the resulting quantum states by tailoring the confinement potential with atomic-scale precision. Using scanning tunnelling microscopy and spectroscopy, we pattern individual Cs ions into designer quantum dot structures on the surface of indium antimonide, in which electrons from a two-dimensional electron gas are confined with chosen in-plane electric-field gradients. We then quantify the atomic level structure, both spatially resolving the orbital character of the electronic states and their magnetic-field evolution. We demonstrate that the level structure, including the induced zero-field splitting, can be tailored by the designed geometry of the local electric fields. These effects can be described using a Hamiltonian that allows consistent treatment of the confinement-induced spin-orbit coupling beyond the conventional Bychkov-Rashba description. This Hamiltonian is derived from a multiband k.p model and takes the energy dependence of the relevant physical parameters into account. Such precise control of spin-orbit coupling in semiconductor quantum dots is relevant to quantum and spintronic technologies.

The stationary measure of equilibrium systems with detailed balance follows a Boltzmann distribution, so that for short-ranged interactions the measure is local, meaning that distant spatial domains are statistically independent. In contrast, active systems break detailed balance, and can have nonlocal stationary measure even for fully local dynamics. Here, by expanding in nonlinearity about a Gaussian-model limit, we construct the measure perturbatively deep in the disordered phase for a class of models that includes Active Model A, Active Model B+, Model AB, the Nonreciprocal Cahn--Hilliard model, and the Toner--Tu model. In this regime, nonlocality is linked to a dynamical conservation law. Our results generically preclude construction of a Landau--Ginzburg expansion of the stationary measure (as opposed to the dynamical equations) for conserved active field theories.

In dual-gated bilayer graphene, a perpendicular displacement field opens a band gap that modifies both the channel conductance and the electronic compressibility, motivating measurements that resolve resistive and capacitive responses on the same device. We integrate an hBN-encapsulated bilayer graphene heterostructure with an on-chip superconducting Nb lumped-element LC resonator and carry out RF reflectometry near 4.25 GHz. DC transport and finite-bias spectroscopy on the same device provide a transport reference. Top and bottom gates independently set the carrier density and displacement field. The DC and RF gate maps share the same gate-dependent features, with finite-bias measurements revealing a region of suppressed conductance whose bias extent grows with the displacement field, consistent with a field-induced gap. The gate-dependent resonance-frequency shift is converted to the effective capacitance seen by the resonator using an equivalent-circuit model. The capacitance shows a minimum near the conductance-suppressed region, consistent with reduced electronic compressibility in the gapped bilayer graphene, and exhibits an electron-hole asymmetry. The on-chip configuration probes the gate-dependent admittance of a dual-gated van der Waals heterostructure, providing capacitance-sensitive information that complements DC transport within a single device.

We investigate the optical response of Kekulé-Y graphene under uniaxial strain and carrier doping. Using a low-energy effective Hamiltonian, we show that strain reshapes the low-energy electronic structure of the Kekulé-Y phase and induces Van Hove singularities at energies well below those of pristine graphene. Within the Kubo formalism, we calculate the optical conductivity and identify multiple anisotropic interband features, with a pronounced resonance arising from strain-induced Van Hove singularities. The pronounced resonance is strongly anisotropic and robust against moderate thermal broadening and disorder, providing a clear optical signature of Kekulé-Y ordering. We further derive analytical expressions for the low-energy optical conductivity and the Drude weight, providing a detailed characterization of the strain- and doping-dependent optical response. Our results establish strain engineering as an effective route for controlling valley-dependent optical properties in Kekulé-Y graphene, originating from the Kekulé-induced coupling of the Dirac valleys, and suggest feasible optical probes for the experimental identification of the Kekulé-Y phase.

Memorization in diffusion models is often treated as a global property of the model or dataset. In practice, however, a single diffusion model can simultaneously generate both memorized and novel samples. Which training samples are most likely to be memorized? In this work, we show that memorization is governed by \emph{local data coverage}. Leveraging the connection between diffusion models and kernel density estimation (KDE), we derive a theoretical criterion that predicts whether a point is memorized based on the density of training data in its neighborhood and the size of the training dataset. In the high-dimensional limit, this leads to a sharp, local transition: regions of low coverage are dominated by isolated training samples, which are memorized, while dense regions support interpolation and generalization. We validate these predictions empirically, showing that memorization increases with local sparsity and that diffusion models exhibit a coexistence of memorized and novel samples within the same model. Extending this framework to multi-class settings, we further show that classes with higher intra-class sparsity (and thus lower local coverage) are more strongly memorized. Our results provide a local view of memorization in diffusion models, explaining when and where memorization occurs in terms of data geometry.

We develop classical density functional learning to address fluids with truncated pairwise interparticle interactions in three-dimensional spherical geometry. Simulation data for systems with randomized repulsive pair potentials provide the basis for supervised training of a neural metadensity functional, thereby making efficient use of results for radial distribution functions in the bulk fluid via the test particle route. Specifically, we develop spherical local learning in order to represent the one-body direct correlation functional in terms of a neural network, which captures spatial curvature effects as well as the metadensity functional dependence on the thermally scaled pair potential. The framework yields efficient access to inhomogeneous structuring and related physical phenomena that occur in fluids and general solvents when adsorbed against curved solutes and confined inside of spherical and planar cavities. Test particle setups facilitate accurate prediction of the bulk fluid pair structure and verification of thermodynamic test particle sum rules via functional line integration. Applying the metadensity functional for Henderson inversion allows one to infer accurately the pair potential from the bulk radial distribution function. We address implications of the geometrical setup for two-body quantities and obtain the two-body direct correlation functional from automatic differentiation. For the hard sphere fluid, we confirm metadensity functional predictions against results from a standard neural density functional with fixed pair potential as well as to an analytic functional as given by fundamental measure theory. Simulation results provide further reference and corroborate reliable results of the spherical neural metadensity functional across a broad range of applications.

At the intersection of rising wealth inequality and intensifying environmental pressures, we investigate a reverse causal relationship that has received comparatively little attention: wealth inequality may not only be a consequence of environmental crises, but also act as a structural obstacle to the ecological transition itself. We develop a stylized agent-based model in which heterogeneous agents, whose initial wealth follows a Pareto distribution, allocate their income between either a Brown or a Green sector through a utility function. The function is designed to capture the trade-off between short-term returns and exposure to long-term systemic risks. A central ingredient is that wealthier agents perceive themselves as less vulnerable to environmental shocks, thereby reducing the amount of resources available for the transition. We show that, beyond inequality thresholds compatible with those observed in most developed countries, the economy remains locked in a Brown regime, even when a substantial share of agents is sensitive to externalities. We then assess a set of stylized fiscal policies (basic income, carbon taxation, Green incentives, and a combined scheme) and find that their effectiveness depends strongly on the inequality regime and on the regressivity embedded in the fiscal mechanism, revealing multidimensional trade-offs between transition speed, cumulative environmental destruction, growth, and fiscal pressure.

The dissipation in the relaxation of an ensemble of conservative stochastic systems towards the steady state is quantified by the free energy difference. Functional inequalities within the framework of [F. Otto and C. Villani, J. Funct. Anal. 173, 361 (2000)] are here revisited which connect the free energy dynamics and optimal transport, offering a geometric perspective on the instantaneous speed of relaxation in the presence of potential barriers. These are illustrated with numerical relaxation experiments on Landau-Ginzburg potentials.

We investigate the linear and nonlinear rheology of dense chiral fluids composed of self-spinning particles under external shear. Using particle-based simulations of a two-dimensional Lennard-Jones model with transverse interactions, we show that chirality acts as an intrinsic source of fluctuations and shear. In the solid regime, spinning fluidizes the system, weakening hexatic order. In the liquid regime, the viscosity is quantitatively described by a Green-Kubo relation upon replacing the temperature by a chirality-dependent effective temperature. Beyond linear response, flow curves collapse when expressed in terms of the ratio between imposed shear and spinning rates, revealing a thinning-by-spinning mechanism. At large forcing, this correspondence breaks down and a pronounced handedness asymmetry emerges: when transverse interactions oppose the imposed shear, stresses relax through the formation of string-like flow channels. Our results identify chirality as a generic mechanism for fluidization and provide a unified framework for understanding the rheology of dense chiral suspensions.

We develop a thermodynamic framework for non-equilibrium affinities based on generalized entropies. In particular, we extend the classical concept of De Donder by introducing $q$-affinities associated with Rényi and Tsallis entropies. This in turn allows us to generalize thermodynamic driving forces to systems with long-range interactions and/or strong correlations. For Rényi entropy, we build on a thermodynamic interpretation due to Baez, where the entropy is expressed through finite differences of the Helmholtz free energy at two temperatures. This leads to a generalized thermodynamic potential whose derivative with respect to a reaction coordinate defines the Rényi $q$-affinity. The resulting expression admits a representation in terms of exponential work averages, establishing a connection to Jarzynski-type fluctuation relations. For Tsallis entropy, we consider Markov jump processes using a master-equation-based approach. We derive a $q$-deformed entropy balance law and obtain an explicit expression for the Tsallis entropy production rate, proving its non-negativity and thus recovering a generalized second-law structure. This allows to identify a local stochastic $q$-affinity with the generalized thermodynamic force entering the entropy production rate.

Physical vapor deposition (PVD) provides a route to prepare highly stable and anisotropic organic glasses that are utilized in multi-layer structures such as organic light-emitting devices. While previous work has demonstrated that anisotropic glasses with uniaxial symmetry can be prepared by PVD, here, we prepare biaxially aligned glasses in which molecular orientation has a preferred in-plane direction. With the collective effect of the surface equilibration mechanism and template growth on an aligned substrate, macroscopic biaxial alignment is achieved in depositions as much as 180 K below the clearing point $T_{LC-iso}$ (and 50 K below the glass transition temperature $T_g$ ) with single-component disk-like (phenanthroperylene ester) and rod-like (itraconazole) mesogens. The preparation of biaxially aligned organic semiconductors adds a new dimension of structural control for vapor-deposited glasses and may enable polarized emission and in-plane control of charge mobility.

Pseudo entropy is a complex-valued generalization of entanglement entropy defined from a reduced transition matrix. We study the pseudo entropy associated with a real-time transition matrix between an initial pure state and its unitary time evolution. For a subsystem $A$, we show that the short-time behavior of real-time pseudo entropy is governed by the correlation between the physical Hamiltonian $H$ and the modular Hamiltonian $K_A=-\logρ_A$ of the initial reduced state, $ S_A(t,0)=S_A(0)-it \langle K_A(H-\langle H\rangle)\rangle + \mathcal{O}(t^2)$. For Hermitian dynamics, the initial imaginary response is controlled by the symmetrized covariance of $H$ and $K_A$ with an overall minus sign, while the initial real response is governed by their commutator. Thus the imaginary part of real-time pseudo entropy is not merely a branch artifact: it is a time-oriented modular response generated by the correlation between microscopic time evolution and subsystem coarse graining. We clarify the relation of this result to the known first law of pseudo entropy, derive an all-order expression in a Schmidt-diagonal model, recover thermal pseudo entropy as a special case, illustrate the covariance/commutator decomposition in a two-qubit model, and confirm the covariance response in transverse-field Ising-chain quenches, including a finite-size study of a modular susceptibility near the Ising critical region. We discuss how this amplitude-level oriented response can be related to ordinary entropy production, and also give a concrete $\mathcal{PT}$-symmetric toy-model illustration of the non-Hermitian extension.

We construct stochastic thermodynamics of overdamped Langevin systems on nonrelaticvistic curved spaces with time-dependent metrics. The time dependence of the metric contributes to the energy balance by performing work on the kinetic energy, which is instantaneously dissipated as heat in the overdamped regime. This contribution makes our framework thermodynamically consistent so that entropy production satisfies the second law of thermodynamics. As a special case, when the metric evolves according to backward Ricci flow, the entropy balance exhibits a structure similar to Perelman's entropy functional. Our framework provides a way to quantify thermodynamic costs in dynamics on time-evolving spaces such as diffusion on membranes.

In distributed quantum networks, interacting spin systems can mediate the generation of highly entangled links between distant nodes. We investigate the role of effective parity-time (PT)-symmetric non-Hermitian spin-1/2 bulks weakly coupled to two quantum links, obtained due to the environmental interactions affecting both the bulk and the links. Focusing on effective non-Hermitian nearest-neighbor (NN) Su-Schrieffer-Heeger (SSH) models, we analyze how non-Hermiticity influences the dynamical formation of long-distance entanglement (LDE). For a paradigmatic model consisting of a quantum XX bulk subjected to imaginary staggered magnetic fields, we analytically determine the exceptional points arising from the resulting bulk-mediated interactions between the links. Combining analytical and numerical methods, we demonstrate that an initially fully separable state can dynamically evolve into highly entangled link states near these exceptional points in the broken regime. Further, after optimizing over time and system parameters, near-unit time-averaged entanglement between the links emerges under weak imaginary magnetic fields and bulk-link couplings, which cannot be attained in the corresponding Hermitian systems. Moreover, the non-Hermitian dynamics exhibit a freezing of high entanglement in the vicinity of exceptional points, a feature absent in Hermitian counterparts. We also identify regimes of long-range interaction strengths that yield a higher time-averaged entanglement than the corresponding NN models. Furthermore, we establish that LDE persists in the stationary regime, highlighting the promise of engineered non-Hermitian dynamics for realizing robust and frozen entangled links in quantum networks.

A scalar potential landscape is one of the most useful ways to understand the stability and transition of a dynamical system. For non-gradient dynamics, however, the construction of a global Lyapunov-type scalar for nonlinear flows with recurrent structures remains a major obstacle. We introduce the generalized Lyapunov function, a scalar function that is non-increasing along deterministic trajectories, as a unifying notion of nonequilibrium potential. Ordinary Lyapunov functions, Freidlin--Wentzell quasi-potentials, and Ao-type potentials are recovered as special representatives. We then propose a data-free physics-informed neural-network framework in which the Lyapunov inequality and a weak divergence-scale compatibility condition are directly embedded into the loss function. The method is tested on linear systems, the Hopf normal form, the van der Pol oscillator, and a three-dimensional Hopf-link flow with two linked limit cycles. The learned landscapes agree with available analytical benchmarks and reveal the invariant sets as low-potential or constant-potential structures, providing a practical route to potential-landscape construction for nonlinear non-gradient systems.

We investigate the quantum entanglement of Bethe states across a family of integrable spin chains, including the XXX$_{\frac{1}{2}}$ model, its higher-spin generalizations (XXX$_s$), and the non-compact $SL(2,\mathbb{R})$ chain. For on-shell eigenstates, we perform a comprehensive scan of the bipartite entanglement entropy across the entire spectrum of finite chains with periodic boundary conditions, and identify the Bethe solutions that minimize and maximize the entanglement. These extremal solutions follow systematic, spin-dependent patterns in the Bethe quantum numbers. In the XXX$_{\frac{1}{2}}$ spin chain, for the antiferromagnetic chain, the state with minimal entropy always coincides with the lowest-energy state (the ground state) within a given fixed-magnon sector. For the higher-spin XXX$_s$ model, however, the lowest-entropy state is not always identical to the ground state, and can even be the state of highest energy. By contrast, the Bethe roots that maximize entropy exhibit considerably more intricate structure. Our analysis further reveals how special Bethe root configurations, such as singular and strange solutions, affect entanglement, and it uncovers characteristic entanglement features in the non-compact $SL(2,\mathbb{R})$ chain that are absent from compact spin chains. For off-shell Bethe states, we develop an optimization algorithm that extremizes the entanglement entropy over rapidity distributions, enabling us to explore the maximum entanglement achievable by a Bethe state without imposing the Bethe ansatz equations.

We study the collective dynamics of independent particles that diffuse outside a spherical surface, on which they are replicated with a prescribed catalytic rate. In spatial dimensions three and higher, the transient nature of diffusion creates the competition between autocatalytic and escape events, thus leading to a rich phase diagram between subcritical (extinction), critical (steady-state), and supercritical (growth) regimes at long times. The rotational symmetry of the domain and an explicit form of the single-particle diffusion propagator allow us to obtain the statistics of the population size (i.e., the number of particles). In this way, we analyze the mean population size, its variance and higher-order moments, as well as the full distribution. In particular, we obtain a fully explicit form of the distribution at long times and describe a slow, power-law approach to this steady-state limit.

How a two-dimensional solid melts remains unsettled after 60 years of study, as theory, model systems, simulations, and atomic-resolution experiments continue to suggest conflicting scenarios. The same transition can appear continuous or abrupt depending on how it is observed, where this ambiguity is especially acute in confined water. Here we study bilayer water under nanoconfinement and ask not only where its phase boundaries lie, but how the system responds to the two fields that drive them: temperature and lateral pressure. Using Lee-Yang zeros together with enhanced sampling, we find that some phase boundaries are field-selective: the two responses can differ either in continuity itself, or in how strongly they are rounded in finite systems. This distinction changes the two-step melting picture. The solid--hexatic transition is field-selective first-order, with the density channel remaining unusually rounded, whereas the hexatic--liquid transition becomes a conventional first-order transition once larger cells reveal a hidden bimodal enthalpy distribution. This framework organizes the apparent disagreement among confined-water simulations, hard-disk models and AgI experiments by identifying which thermodynamic channel each probe sees.

Mechanical computation, in which logic functions are realized through deformation rather than electronics, has been demonstrated in systems such as origami, kirigami, and mechanical metamaterials. In these systems, logic states and functions are typically determined by geometry and material properties, making it sensitive to deformation and imperfections. Here we introduce a mechanical computing architecture in which logic is defined by topology rather than geometry. The circuit is realized as a knitted textile formed from a single continuous yarn, where information is encoded in the topology of stitches and processed through controlled unraveling. By discretizing the textile into a lattice of interacting cells, we construct topological propagation rules that implement universal logic operations, including NOT, AND, and OR gates, as well as a half-adder. Experiments demonstrate that the logical output is robust against geometric deformation, while mechanical factors affect only if the computation can be executed. These results establish topology-defined computation as a model for information processing in textiles and other reconfigurable physical systems.

Myopic self-avoiding fractional Brownian motion (FBM) is a stochastic process in which an ensemble of particles is driven by fractional Gaussian noise while being repelled by the gradient of the time-integrated ensemble density [J. House, R. Bakhshizada, S. Janušonis, R. Metzler, and T. Vojta, Phys. Rev. E 112, 034119 (2025)]. Depending on the anomalous diffusion exponent $α$ characterizing the noise, the process features two dynamical regimes: an interaction-dominated regime ($α< α_c=4/(d+2)$) where the mean-density interaction governs long-time dynamics, and a noise-dominated regime ($α> α_c$) where FBM correlations prevail. In the interaction-dominated regime, the mean-squared displacement grows as $\langle r^2(t) \rangle \sim t^{4/(d+2)}$ regardless of $α$, while for $α> α_c$ the standard FBM scaling $\langle r^2(t) \rangle \sim t^α$ is recovered. Here, we develop an analytical mean-field theory of myopic self-avoiding FBM, based on a Fokker-Planck approach to the interaction-dominated regime. This allows us to derive closed-form polynomial solutions for the probability density. To compare with computer simulations, we develop an efficient radial binning algorithm that significantly reduces the computational complexity, making large-scale three-dimensional simulations feasible. Extensive simulations in one, two, and three dimensions confirm the analytical predictions. We also discuss the application of the process to the self-organization of serotonergic axons (fibers) in vertebrate brains, where FBM paths with self-avoidance provide a natural framework for understanding spatial heterogeneities of fiber densities.

We study the mean first-passage time of a random walker to a small absorbing target at the center of a two-dimensional annulus with a specularly reflecting outer boundary. The problem is motivated by natural killer cell migration toward a target cancer cell, where the goal is to quantify how long it takes immune cells to reach the target and how search efficiency depends on directional persistence and chemotactic bias. Cell motion is modeled as a velocity-jump process. We first consider a correlated random walk with a von Mises turning kernel, with a concentration parameter controlling directional persistence. We then extend the model to a biased correlated random walk using a phase-shifted turning kernel that represents preferential motion, for example following a concentration gradient. Our analysis combines closed-form benchmarks for simple and biased random walks, Fourier-mode reductions of the transport equations for the correlated and biased correlated models, and a fast-turning perturbation expansion that gives an analytical correction to the diffusion-limit mean first-passage time for the random walker. Our analytical results are supported by numerical methods that include a semi-Lagrangian solver in radial and angular coordinates, a stationary discretisation designed to handle biased transport, and an event-driven Monte Carlo simulator for cross-validation. Together, our results provide a framework relating persistent and biased immune-cell motion to target-search times in confined two-dimensional domains.

We report the structural, electronic, and transport properties of the goldene 2D material using the density functional tight-binding (DFTB) method. Electronic transport calculations were conducted in conjunction with the non-equilibrium Green's functions (NEGF) technique. Our study reveals that the Au 2D material is dynamically and thermally stable, and it possesses good elastic properties. On the other hand, goldene has a linear relationship between current and voltage at low potentials, indicating its metallic character. The calculated current-potential curve correlates well with transmission functions and the electronic density of states around the Fermi level. We also investigated the electronic structure and magnetic properties of silicon (Si)-doped Au 2D material. Our results show that the Si atom can induce a local magnetic state in the goldene monolayer. The resulting magnetic moment is 0.63 $μ_B$.

Many-body Floquet models with exact space--time symmetry, such as the kicked Ising spin chain (KIC), provide natural examples of systems with dual-unitary dynamics. The requirement of exact space--time symmetry is, however, highly restrictive, as it permits only nearest-neighbor interactions. Based on a pair of Hadamard matrices, we construct a wide family of dual-unitary kicked spin chains with long-range interactions. We show that local two-point correlations in such models propagate along the light-cone edges \( |n| = r|t| \), where \(r\) is the interaction range, and can be derived analytically for operators with local support. This approach is illustrated using the example of a kicked Ising spin chain with next-to-next-neighbor interactions.

We present a direct collective-spin derivation of the effective uniform-mode Hamiltonian used in cavity magnonics. Starting from a nearest-neighbor Heisenberg ferromagnet coupled to long-wavelength magnetic fields, we show that the relevant dynamics can be restricted to the fully symmetric spin sector, where the exchange interaction contributes only a constant energy shift and the ferromagnet behaves as a macrospin of length $Ns$. Applying the Holstein--Primakoff transformation directly to this total spin yields the usual uniform magnon mode and its leading nonlinear corrections without first introducing site-resolved bosonic operators. This collective formulation makes explicit the interpretation of the ferromagnet as a synthetic large-spin atom and provides a compact route to the effective Hamiltonians used in driven and Floquet cavity magnonics. As a physical consequence, the leading nonlinear correction produces an occupation-dependent reduction of the effective magnon--photon coupling, providing a simple signature of finite-spin saturation under strong uniform-mode driving.

Stabilizer Rényi entropy has emerged as a probe of nonstabilizerness in quantum many-body systems, but its scaling structure beyond critical points remains poorly understood compared with entanglement entropy. Recent field-theory approaches indicate that stabilizer entropy contains universal critical data and boundary-sensitive terms, raising the question of how these structures extend into massive and crossover regimes. We address this problem for a broad class of finite-range spin chains at Rényi index one-half. We derive exact finite-size formulas for both full periodic chains and finite intervals of the infinite chain, making the universal crossover from critical to noncritical behavior analytically accessible. In periodic geometry, the entropy obeys a volume law away from criticality and exhibits a universal finite-size crossover controlled by the competition between system size and correlation length. We also show that the large-scale SRE density develops a cusp across the field-tuned critical line, while the XX endpoint is governed by a distinct scaling regime associated with the saturation point. In the subsystem geometry, the interval entropy separates bulk critical behavior from boundary contributions generated by the way the finite region cuts the infinite chain. The crossover from critical to massive behavior is then encoded in boundary constants and universal functions controlled by the correlation length. Through exact stabilizer-entropy correspondences, the scaling theory extends to internal XY reductions, Finite-range spin chains, and Cluster--Ising representatives. Our results provide an exact lattice benchmark for the emerging QFT description of stabilizer entropy beyond isolated conformal points.

We present a detailed study on how the strength of the interlayer magnetic coupling on Fe$_{85}$Co$_{15}$/Ni$_{80}$Fe$_{20}$ bilayers modifies the spin wave behavior of this system. A series of Fe$_{85}$Co$_{15}$/Ni$_{80}$Fe$_{20}$ bilayers deposited on MgO[100] substrates were grown by magnetron sputtering. Magnetic characterization of the samples was performed using a vibrating sample magnetometer and magneto-optical Kerr effect. The in-plane hysteresis loops reveal a cubic magnetic anisotropy of magnetocrystalline origin, with easy and hard axis along the [100] and [110] Fe-Co crystallographic directions, respectively. Ferromagnetic resonance measurements were performed to analyze the in-plane angular dependence of the resonance field, and also the resonance field at several frequencies was determined along the hard axis. By using a bilayer model in the frame of the Landau-Lifshitz-Gilbert magnetization equation of motion, the magnetization precession components were calculated, as well as the dependence of precession area on the Fe-Co layer thickness and the ferromagnetic interlayer coupling. We observe a maximum in the area of the ellipsoid generated by the magnetization precession of the permalloy layer at a certain exchange constant, showing that this effect could be used to maximize the injected spin currents, which could be tuned by changing the interlayer exchange constant in bilayer systems, the saturation magnetization of the materials, or the excitation frequency.

We study the thermal transition between the bounce and the sphaleron in quantum mechanics with a metastable vacuum from the viewpoint of Borel resurgence. For two models representing a second-order and a first-order transition, we compute the perturbative expansion of the thermal free energy to high orders and extract the leading Borel singularity data $(A,b,S)$ as functions of temperature. The Borel singularity location $A$ reproduces the on-shell action of the dominant saddle on both sides of the transition, joining smoothly in the second-order case and developing a kink in the first-order case. The characteristic exponent $b$ jumps between $0$ and $1/2$ across the transition, counting the zero modes of the corresponding saddle. The Stokes constant $S$ matches the one-loop determinant around the saddle. The perturbative expansion around the false vacuum thus determines the transition temperature, the order of the transition, and the decay rate including the one-loop prefactor without relying on semiclassical inputs.

Nonvolatile gate-driven switching of quantum anomalous Hall (QAH) states in graphene moiré systems provides a promising route toward topological electronics based on chiral edge states. However, deliberate use of this switching mechanism requires control over both the magnetic properties and metastability of QAH states. While previous demonstrations mostly relied on the intrinsic magnetic energy landscape of moiré devices, here we show that this landscape can be engineered through proximity coupling to WSe2. We find that proximitizing twisted monolayer-bilayer graphene by WSe2 reshapes the magnetization reversals responsible for nonvolatile electrical switching of QAH states. We attribute this effect to the proximity-induced spin-orbit coupling (SOC), which can lock spin and valley and modify the magnetization of the competing states involved in switching compared with non-proximitized graphene systems. Our findings establish proximity-induced SOC as a new way to engineer magnetic properties and switchable magnetic states in graphene-based systems. We further demonstrate that strong magnetic metastability in tMBG allows the magnetic states to be gate-tuned between QAH and metallic regimes, and between QAH states with Chern numbers |C| = 2 and 1 without resetting the magnetic state. This functionality points toward new device architectures based on QAH chiral edge states.

Stochastic dynamics with state-dependent diffusion are widely used for Brownian motion in confined, anisotropic, and hydrodynamically coupled systems. The conventional Langevin formulation includes a spurious drift associated with multiplicative noise, but its free energy generally does not transform as a scalar, meaning that the covariance is not explicit. Here, we formulate a geometrically consistent Langevin equation by introducing a scalar free energy and using the diffusion tensor as a metric on configuration space. The spurious drift is then expressed as a Christoffel contribution of the diffusion metric. While our formulation is equivalent to the conventional one through the relation between the non-scalar and scalar free energies, it makes the coordinate covariance explicit. We demonstrate its consistency in representative examples of state-dependent diffusion arising from coordinate transformations, geometrical confinement, and projection from curved to flat spaces.

Symmetry is a powerful way to control coupling between photonic mode families. In spatially periodic structures, glide symmetry can protect band contacts and suppress stop bands. Here we show a different role for temporal glide, a spatiotemporal counterpart combining reflection with a half-period time translation. In a scalar time-modulated trilayer waveguide, temporal glide imposes an exact selection rule linking frequency conversion to transverse-mode symmetry: the parity content of every Floquet eigenstate alternates with sideband index, up to a state-dependent sign. In scattering, this means that a mode of definite parity can emit only into the opposite transverse parity at odd sidebands and into the same parity at even sidebands. We verify the rule directly in bulk Floquet eigenstates and in finite-section time-domain simulations. An incident odd waveguide mode is converted into an even frequency sideband, while all symmetry-forbidden output channels at the analysed sidebands are suppressed to numerically negligible values. Rather than acting as a temporal copy of spatial-glide band sticking, temporal glide provides a distinct symmetry principle for converting electromagnetic energy between selected modes and frequencies.

Digital Alchemy (DA), introduced by Van Anders et al., is a statistical mechanics-based generalized thermodynamic ensemble method that employs computer simulations to optimize colloidal particle design. This approach applies the principles of statistical mechanics to predict and tailor particle attributes that lead to desired self-assembled structures or material properties. However, as an inverse design method, its main limitation is that the target structure must be known \textit{a priori}. Therefore, the optimal design from DA does not guarantee the targeted structure is the most or the only stable one. This highlights the importance of forward design with an exploratory scheme for optimizing novel colloid designs, which becomes more suitable in such cases. In this paper, we introduce Exploratory Digital Alchemy (EDA), an enhanced forward design scheme that begins by releasing the constraint of the target crystal from DA, followed by an exploration-oriented bias that has been extensively used in enhanced sampling methods such as metadynamics (MetaD). We demonstrate the utility of EDA through examples involving particles interacting via a two-dimensional Lennard-Jones Gauss potential (LJGP) and a three-dimensional oscillating pair potential (OPP). We applied EDA to study the free energy landscapes given different potential parameters of LJGP at different temperatures. With the exploratory scheme, we've also successfully identified a wide range of OPP potential parameters that stabilize metastable Frank-Kasper phases. Our approach fuses the standard DA framework with metadynamics, which could potentially be useful for studying alchemical reactions in a generalized ensemble.

Mechanical characterization of soft materials at high strain rates is challenging due to their high compliance, nonlinear viscoelastic behavior, and potentially history-dependent responses. Inertial microcavitation rheometry (IMR) addresses this challenge by coupling laser-induced cavitation (LIC) experiments with numerical simulations of bubble dynamics models to infer constitutive models and material parameters. Both IMR and its variants infer parameters that depend on the chosen fitting window, which suggests that a constant-parameter constitutive model is insufficient to describe the full cavitation event. We use this window dependence to identify when the constant-parameter assumption fails, rather than to report a single effective parameter set. The constitutive parameters are estimated over moving, overlapping windows using a modified iterative ensemble Kalman smoother with multiple data assimilation (MIEnKS-MDA). Within the neo-Hookean Kelvin--Voigt (NHKV) constitutive model, we obtain time-resolved estimates of the constitutive response in polyacrylamide (PAAm) hydrogels with different crosslinker concentrations. The inferred shear modulus and viscosity generally decrease and then plateau during cavitation, while exhibiting relatively weak temperature sensitivity. For gelatin gels, by contrast, the inferred property evolution shows a pronounced temperature dependence, with distinct trends at low and high temperatures. Moreover, both the apparent shear modulus and viscosity exhibit significant variations during the first two bubble collapses. These results show that time-resolved parameter estimation within the prescribed NHKV constitutive structure can diagnose where the constant-parameter model assumption falls short during cavitation, thereby guiding the development of improved physics-based models of complex bubble--material interactions.

The development of miniaturized non-reciprocal microwave technologies compatible with integrated circuit architectures remains a critical challenge for modern information technology. Here, we present the first experimental characterization of a micron-sized prototypical magnon circulator. Taking advantage of the chiral excitation of spin-waves via nanowire gratings, we propose an original design of a circulator involving three channels of rectilinear and unidirectional spin-wave beams. We demonstrate via a full 3-port spin-wave spectroscopy a genuine spin-wave circulation between the three ports. The narrow frequency band of operation can be tuned over a broad range of frequencies ($2$-$8$ GHz) with both an external field of up to $100$ mT, and the dimensions of the grating specifying the wavevectors. This proposed scheme opens up possibilities for new architectures of integrated and miniaturized non-reciprocal microwave devices.

We study a three-terminal thermoelectric engine, focusing on the role of quantum coherence and information flow. A double-dot connects two reservoirs at different chemical potentials, while a third dot monitors their occupation via Coulomb interaction and can be interpreted as an autonomous Maxwell demon. Within the parameter range where the device operates as an engine, we identify conditions under which this interpretation holds. The system dynamics is described within a Redfield master equation that allows us to identify two distinct dynamical regimes with steady states well captured by suitable Lindblad approximations. These two regimes define a first quantum-to-classical transition controlled by the interdot tunneling strength. We further consider the effect of a phonon bath coupled to the double-dot, which induces a second quantum-to-classical transition by generating incoherent transport and decoherence in the small interdot tunneling regime. We identify a parameter region where phonon-induced decoherence suppresses both the coherent transport contribution and the information flow toward the monitoring dot, suggesting that coherence can enhance the demon mechanism in this regime. By tracking information and transport properties across these crossovers, our model shows how coherent tunneling, decoherence, and incoherent phonon-assisted transport compete in an autonomous information engine, while clarifying which thermodynamic Lindblad description is appropriate in each regime.

We investigate the population dynamics of surface-mediated autocatalytic processes, in which particles diffuse in a complex environment towards surface regions where they can be either killed or replicated. These opposite mechanisms compete with each other and lead to a sophisticated stochastic evolution of the population size. We provide a systematic analysis of the generating function of the population size. We also deduce its distribution, mean, variance and higher-order moments. For this purpose, we employ several equivalent descriptions of these quantities in terms of nonlinear integral equations and partial differential equations with nonlinear boundary conditions. We inspect the long-time behavior of the population dynamics in three regimes when the mean population size vanishes, reaches a steady-state level, or grows exponentially. A numerical solution of the underlying integral equations and independent Monte Carlo simulations support our theoretical predictions.

We study the interference of critical dynamics associated with zero modes (ICDZM) in the generalized Creutz ladders using closed quench paths that pass through two critical points successively. By reading out the final zero-mode transfer probability, we find rich ICDZM interference patterns dependent on the quench path. In particular, when the closed path links two topologically nontrivial phases, the ICDZM pattern may either vanish or exhibit period doubling. Within the framework of WKB analysis, this phenomenon is well clarified by the interference phase accumulated in the quench procedure. We also demonstrate that the zero-mode transfer probability can be detected by the deviation of the boundary particle number from its initial fractional value, which arises from the blending of bulk modes in the critical dynamics. As an edge defect, the zero-mode transfer probability captures both the ICDZM oscillation and the known anomalous defect production in a non-closed quench path. These results identify ICDZM and the corresponding edge defect as probes for critical dynamics associated with topological zero modes.

Topological defects govern much of the flow behavior and orientational order in active nematics, making their control relevant for active matter physics, smart materials, and microfluidics. Applied activity patterns can induce self-propulsion of active nematic defects, but general-purpose methods for exploiting this effect to control defects remain largely unexplored. Here we use deep reinforcement learning (RL) to perform minimum-time position control of +1/2 defects in hybrid lattice Boltzmann simulations of active nematodynamics. Spatiotemporally patterned activity, implemented as a control field in the active stress, steers defects through microchannel geometries and reveals finite-time reachable regions of defect position space. Reachability is shaped by director anisotropy, homeotropic wall anchoring, and the allowed activity patterns: local patterns steer defects in free domains but fail in junctions, whereas global patterns open otherwise inaccessible channels. In constrained geometries, the original defect may be unable to reach some goals intact, but controlled pair creation enlarges the effective reachable set by transferring control to a newly created +1/2 defect. The trained RL controllers outperform static and rule-based baselines, and controllers trained only on simple junctions can be combined without fine-tuning into a meta-controller that successfully steers defects through a larger test maze. Free energy visualizations show that guided defects write persistent, history-dependent distortions into the director field that can later be partially erased by -1/2 defects. Thus, RL-based control uncovers how confinement, anchoring, actuation geometry, and defect creation determine reachable motion in active nematics, providing a framework for other control tasks in soft and active matter.

Extensive systems have a simple thermodynamic signature: the logarithm of the partition function scales homogeneously with the size of the system. We show that the failure of this scaling, measured by the replica energy ${\cal E}$, provides a useful bridge between statistical mechanics and quantum field theory. The associated differential operator $(1-\frac1d L\partial_L)$ removes the leading bulk contribution to $W=\log Z$ and isolates the part that is sensitive to boundaries, topology, defects, long-range forces, or other sources of nonadditivity. In quantum field theory this thermodynamic idea has two closely related uses. For ordinary finite-volume or spherical partition functions, suitable higher-order versions of the same filter remove local counterterms and extract universal fixed-point data such as the central charge, the sphere free energy $F$, and the Euler anomaly coefficient $a$. For replica geometries with entangling defects, the same filtering principle gives the renormalized defect free energy. In $2+1$ dimensions, its $n\to1$ limit is precisely the entropic $F$-function. We use this perspective to distinguish ordinary finite-size corrections, topology-dependent constants in gapped phases, subextensive fracton degeneracies, and genuinely nonextensive systems with long-range interactions such as self-gravitating thermal matter. Replica energy therefore offers a common thermodynamic language for additivity, defect free energies, and renormalization-group irreversibility.

Continuous Attractor Neural Networks (CANNs) traditionally rely on pre-wired recurrent connectivity to model spatial representations, path integration, and anticipatory dynamics. However, the biological mechanisms through which this structured connectivity emerges via learning remain relatively unexplored. This work presents a theoretical framework revealing how continuous attractor connectivity and its computational properties self-organize through Hebbian plasticity, firing-rate adaptation, and global inhibition. We show that translationally invariant inputs naturally drive the emergence of stable, Gaussian-profiled feedforward weights. Crucially, anticipatory dynamics arise spontaneously within these feedforward architectures, shifting the activity bump forward without requiring recurrent excitatory collaterals. This predictive shift can be linearly amplified across multilayer networks, consistent with anticipatory activity observed in the superficial layers of the entorhinal cortex. Furthermore, introducing recurrent interactions allows the network to learn connections capable of self-sustaining a moving bump of activity. Finally, by modulating the network with an external, time-varying baseline current that encodes speed, the system adjusts its intrinsic velocity to function as a precise unidirectional path integrator. Ultimately, this study suggests that prospective coding and path integration are not manually engineered features, but rather naturally co-emergent properties of a single self-organizing competitive network.

We study the gauge invariant dynamically conserved charges, and their corresponding densities, for instanton solutions of Yang-Mills theories in four dimensional Euclidean space, for the gauge group SU(2). Those charges were constructed in [1,2] through the integral equations of Yang-Mills theory, using techniques on generalized loop spaces. We use the integral non-Abelian Gauss law to evaluate the gauge-invariant flux of the magnetic and electric non-abelian fields through spherical surfaces centered at origin of the instanton solution. From such a flux, we define gauge-invariant charge densities by considering the charge within an infinitesimal spherical shell of radius $r$. We discuss the issue of the reparameterization invariance of the charges and densities, and show that the magnetic and electric fluxes for the instanton and anti-instanton, at the Euclidean time $x^4 = 0$ and radius $r=1$, which here corresponds to the size $λ$ of those solutions, are non-zero and observable. Our results give an interesting picture of the internal structure of the instanton, and may be important for the properties of the Yang-Mills $θ$-vacuum.

Internal water waves arise when there is a change in density stratification in a fluid, which may occur in an oceanographical context due to variations in temperature, salinity, or other fluctuations in the equations of state. We present a derivation of nonlinear integrable models for the propagation of interfacial internal waves arising between two fluid layers of different densities (at the so called pycnocline). We examine the integrable Benjamin-Ono (BO) equation as an internal wave model, incorporating underlying currents by permitting a sheared current in both fluid layers. The BO equation arises for a specific small-amplitude asymptotic regime. We show that the BO soliton characteristics are strongly affected by the shear current parameters.

The Lyapunov exponent quantifies the sensitivity of a dynamical system to perturbations, and the full Lyapunov spectrum extends this to every orthogonal direction in tangent space. For cellular automata the spectrum is almost always approximated numerically, and the approximation is delicate. We show that the affine rules, those whose update is a XOR of a subset of the inputs together with a constant, admit an exact Lyapunov spectrum. An affine rule has a configuration-independent Boolean Jacobian, so the spectrum reduces to the logarithms of the singular values of a single constant matrix, with no simulation and no limit involved. Two cases carry a closed form. For an affine cellular automaton on a periodic lattice the Jacobian is a multilevel circulant matrix, and the spectrum is the discrete Fourier transform of the rule's gradient stencil, valid in any spatial dimension. For the parity rule on an arbitrary graph the Jacobian is the adjacency matrix itself, so the Lyapunov spectrum is the logarithm of the absolute adjacency spectrum, and the maximal exponent is the logarithm of the spectral radius. The long-time amplitude of a single-site perturbation then scales with the eigenvector centrality of the seeded node. Reading the periodic lattice as the Cayley graph of an abelian group unifies the two cases. Because they are exact, the affine spectra also serve as benchmarks: they reveal numerical artefacts in previously reported spectra and turn the informal correspondence between spectral radius and dynamical sensitivity into an exact identity.

Distinguishing coherent, turbulent, and chaotic operating regimes in supercontinuum generation is important for understanding nonlinear optical dynamics and optimizing broadband light sources. Experimentally identifying the onset of turbulence remains challenging because the most common metric, first-order coherence, requires access to the complex optical field and cannot be directly obtained from intensity-only measurements. In this work, we investigate whether experimentally accessible statistical observables can identify turbulence in supercontinuum generation. We compare wavelength-integrated variance and kurtosis with simulation-based first-order coherence over a chirp-controlled pulse-duration sweep implemented through additional $β_2$ dispersion. The study combines generalized nonlinear Schrödinger equation simulations with shot-to-shot dispersive Fourier transform measurements validated against optical spectrum analyzer spectra. Statistical intensity distributions were analyzed using histograms, complementary cumulative distribution functions, and kurtosis measurements across the generated supercontinuum bandwidth. Simulations and experiments both revealed heavy-tailed intensity statistics in the intermediate pulse-duration regime associated with reduced spectral coherence. The integrated kurtosis reached a maximum near 600 fs in simulations and near 700 fs in experiments, while the integrated variance within the first 20 dB spectral range decreased with increasing pulse duration. The agreement between simulations and experiments demonstrates that variance- and kurtosis-based observables can serve as experimentally accessible indicators of turbulence in supercontinuum generation. These results show that intensity-only statistical measurements can distinguish coherent and incoherent operating regimes without requiring direct field-resolved coherence measurements.

Multistability is a well-known feature of the inertial Kuramoto system (KMI). Here, we demonstrate that an interplay of phase lag and triadic interactions in KMI leads to distinct hysteresis widths corresponding to different stable states. This phenomenon becomes more pronounced with increasing inertia. Theoretical calculations for the backward branch based on self-consistent analysis show that these multiple widths arise from saddle-node bifurcation occurring at different coupling strengths. Moreover, the forward branch corresponds to oscillatory state and does not admit steady-state solution. The study of multiple hysteresis widths may be useful in modeling power grid systems, information storage, and memory selection in real-word systems.

Partial differential equations (PDEs) regulate the behaviour of countless spatiotemporal systems in the physical and life sciences. In many cases, they encode the coupling between the system's degrees of freedom, leading to nonlinear equations whose solution space is challenging to explore exhaustively. Systematic approaches to PDE model exploration are a holy grail of computational science. In this article, we formulate a criterion for increasing the diversity of a search campaign, based on the PDE residual behaviour under solution deformation. We develop a practical formalism to compute this property and illustrate its role in a few cases of interest.

Pattern formation in reaction-diffusion systems is traditionally analyzed under isothermal assumptions, overlooking the dynamical role of temperature in systems where reactions generate and dissipate heat. Here, we investigate non-isothermal reaction-diffusion dynamics by coupling activator-inhibitor kinetics to a dynamically evolving temperature field that modulates reaction rates through Arrhenius-type dependencies. This coupling introduces an additional feedback mechanism that influences stability and pattern selection. Through analytical and numerical analysis of the Cholrine dioxide-Iodine-Malonic acid (CDIMA) and Schnakenberg models, we demonstrate that thermal feedback modifies dispersion relations by enhancing instability growth rates and shifting pattern selection toward shorter wavelengths. Beyond these intrinsic effects, we identify a boundary-mediated mechanism in which thermal constraints qualitatively alter global dynamics. In particular, fixed-temperature boundaries induce nonstationary behavior in the CDIMA system, whereas the Schnakenberg model exhibits robust stationary patterns. These results establish thermal-kinetic coupling as a general mechanism for controlling pattern formation and highlight the role of boundary-mediated heat exchange as a tunable parameter for spatiotemporal organization.

This work investigates how restitution modeling affects the dynamics of rocking blocks subjected to harmonic excitation. While several studies have reported discrepancies between experimentally observed impact behavior and the predictions obtained using the classical Housner restitution coefficient, the implications of adopting alternative restitution formulations on the global dynamics of rocking systems remain largely unexplored. The system is formulated as a hybrid non-smooth dynamical model and analyzed through bifurcation diagrams, Lyapunov exponents, and basins of attraction for different slenderness ratios. By comparing the classical restitution model proposed by Housner with the alternative formulation of Mao et al., we show that the choice of restitution model strongly influences the predicted system response. The alternative formulation leads to an earlier onset and greater prevalence of complex oscillations, as well as changes in the type, stability, and accessibility of attractors compared to the classical model. However, as the slenderness ratio increases, the dynamical features produced by both formulations progressively converge, indicating a reduced sensitivity to the restitution model for taller blocks. These results provide a dynamical perspective on why alternative restitution formulations, which predict impact responses closer to experimental observations, can produce markedly different behaviors from those obtained using the classical Housner model.

We study the dichotomy between compact satellite and ring formation in proto-planetary disks. Specifically, we examine the behavior of a model system of $N$ identical particles locked into circular, gravitationally-bound orbits around a central body. We treat interactions as dominated by inter-particle collisions, and extract an effective two-particle interaction function based on both theory and simulations. We then demonstrate that the expected dynamics are equivalent to a variant of the Kuramoto model, which undergoes a phase transition as parameters vary. This offers a novel potential explanation for the transition between formation of rings versus moons.

We revisit the problem of scalar dissipation anomaly and scaling of scalar gradients in passive scalar turbulence using theory and data from well-resolved direct numerical simulations (DNS) on grid sizes of up to $8192^3$, spanning Taylor-scale Reynolds numbers $Re_λ=140-1000$ and Schmidt numbers $Sc = 1-512$. The theory is based on a joint multifractal description of longitudinal velocity increments and scalar increments, constrained by Yaglom's law and extended to gradients via a fluctuating Batchelor cutoff scale. The DNS data show that the normalized mean scalar dissipation approaches a single asymptotic value as both $Re_λ$ and $Sc$ increase, although larger $Sc$ requires larger $\re$ to reach this state. In the multifractal framework, this corresponds to an effective scalar Hölder exponent tending to zero, associated with sharp cliff-like scalar fronts, and saturation of inertial-range scaling scalar structure-function exponents. The joint velocity-scalar fractal dimension of the dissipative structures is inferred to approach $7/3$, indicating a non-space-filling support. The framework further predicts that for fixed $Re_λ$, higher-order central moments of scalar gradients are independent of $Sc$. This prediction is confirmed by DNS data and by the collapse of standardized probability distributions of scalar-gradient across Schmidt numbers. These results suggest that the $Sc$-scaling of scalar gradients is dictated solely by scalar dissipation anomaly. In contrast, their $Re_λ$-dependence reflects strong intermittency, which can be directly related to mixed velocity-scalar structure function exponents.

The maturation of quantitative tools for studying the high-level structure of animal behavior, and especially tools which represent spontaneous behavior as a sequence of stereotyped and neurally well-defined 'syllables', demands that the field revisit a fundamental theoretical question: if the coarse structure of behavior can be accurately described by Markov models, what do these models really tell us about behavior? In this work, we explore the theoretical implications of these models and discuss how they allow us to quantitatively formulate questions about the sequence-like nature and effective dimensionality of behavior. One important insight is that the eigenvalues and eigenvectors of various model-associated matrices furnish interpretable time scales and modifications of behavior that occur on those time scales. We illustrate our points using both toy examples and Markov models fit to real data. By analyzing the consequences of Markov representations, we clarify the theoretical meaning of progress in quantifying behavior.

We numerically investigate percolation of independent frictionless glued-sphere rod-like particles under gravity through a disordered static bed of larger spheres. We identify two distinct regimes: a \emph{trapping} regime, where rods stop after percolating a limited distance in the bed and a \emph{passing} regime, where rods percolate continuously with constant mean velocity. The transition between these regimes is governed by the length of the rod and the geometrical trapping threshold for spherical particles based on the rod diameter and the minimum pore throat diameter defined by three touching large spheres. The percolation velocity for all rod geometries, including the single sphere limit, collapses onto a single curve when scaled with the gravitational acceleration and the bed sphere diameter. The results also demonstrate that short rods percolate nearly twice as fast as long rods due to the geometric constraints associated with the disordered pore structure of the static bed. Consequently, long rods are more susceptible to trapping via specific contact configurations with the bed spheres, which differ from those for short rods. These results reveal how shape anisotropy introduces dynamical constraints and thresholds in granular percolation, with implications for predicting segregation in mixtures of non-spherical particles.

Numerous transition-metal oxides have low-temperature antiferromagnetic (AFM) states and high-temperature paramagnetic (PM) phases, where the AFM state is usually insulating while the PM phase can be either insulating or metallic. Without involving strong correlation, we use symmetry-broken density-functional theory (DFT) to obtain the PM phases of insulating NaFeO3 vs the recently discovered metallic NaOsO3. We develop the understanding of insulating and metallic behaviors in paramagnetic oxides by analyzing the interactions between magnetic and positional symmetry breaking: The insulating gap is governed by the competition between the spin disorder that induces a distribution of different magnitudes of local magnetic moments and the polymorphous distribution of off-center atomic displacements. NaFeO3, on the other hand, has large positional displacement with small spin-disorder-induced moments distribution, leading to insulating PM phase, whereas NaOsO3 has a pronounced spin-disorder-induced moments distribution that forces the PM phase to become metallic. Our work identifies this symmetry-breaking competition as a general framework to bridge seemingly disparate metal-insulator behaviors in transition-metal oxides paramagnets without invoking strong correlation.

Magnetic Navigation (MagNav) has emerged as a vital alternative Positioning, Navigation, and Timing (PNT) solution, leveraging Earth's magnetic field for robust navigation in GPS/GNSS-degraded or denied environments. Despite its potential, the successful deployment of MagNav is currently hindered by the lack of standardized, high-fidelity geomagnetic reference maps. Existing datasets, primarily designed for geological exploration or academic research, do not meet the distinct operational requirements of navigation systems regarding spatial resolution, error quantification, and global accessibility. This paper initiates a community-focused dialogue on future geophysical data requirements for MagNav, grounded in extensive real-world flight trials. We distinguish between two primary use cases with divergent data needs: Operational MagNav, which requires globally consistent, queryable, and uncertainty-aware datasets for field deployment, and MagNav R&D, which demands comprehensive access to raw survey data to foster innovation. We provide a prioritized set of recommendations for future data requirements, including the development of cohesive, merged datasets, the inclusion of localized 3D uncertainty estimates, and the expansion of the World Magnetic Model (WMM) core field model to spherical harmonic degree 13 to improve consistency. Finally, we emphasize the strategic necessity of designated test ranges to validate these requirements and ensure the operational robustness of MagNav infrastructure.

Mid-infrared (mid-IR) spectroscopy enables label-free molecular detection and is widely employed in biomedical, environmental, and chemical sensing; however, its broader deployment remains limited by bulky instrumentation and insufficient analytical sensitivity. Photonic metasurfaces supporting strong mid-IR resonances provide a promising route toward compact on-chip spectrometers and enhanced molecular sensing, yet their practical implementation is often constrained by limited spectral coverage and fabrication complexity. Here, we present a scalable, resist-free stencil lithography approach for fabricating arrays of Ge pillar metasurfaces on CaF2 substrates that support polarization-insensitive Mie resonances dominated by electric dipole modes. By varying the geometric design parameters during fabrication, we engineered metasurface arrays with discrete resonances spanning the 950-1700 cm^-1 molecular fingerprint region. Furthermore, leveraging the thermo-optic response of Ge, we achieved dynamic tuning of these discrete resonances, demonstrating continuous and reversible resonance shifts of approximately 36 cm^-1 per metasurface over 300-500 K, corresponding to a tuning rate of 0.18 cm^-1/K. The thermally induced spectral sweeping enables multiplexed detection of poly(methyl methacrylate) vibrational modes and continuous reconstruction of its absorbance spectrum across 1100-1215 cm^-1. These results establish a scalable dielectric metasurface platform with spectrally reconfigurable mid - IR modes for molecular sensing across the fingerprint region and compact infrared sensor technologies.

Raman and infrared anomalies associated with H-bond symmetrization in $δ$-AlOOH, including mode softening and linewidth broadening at 5-10 GPa, occur at significantly lower pressures than predicted by static harmonic theory. To resolve this discrepancy, we combine harmonic phonon calculations with strongly constrained and appropriately normed (SCAN)-based deep-potential molecular dynamics and phonon quasiparticle analysis at 300 K. This framework extracts temperature- and pressure-dependent frequencies and lifetimes from long-time trajectories, capturing the branch reorganization and rapid linewidth growth characteristic of the disordering regime. Incorporating quasiparticle renormalization and directional longitudinal-optical-transverse-optical (LO-TO) splitting further yields near-quantitative agreement with the ambient-pressure OH-stretching Raman multiplet. These results identify finite-temperature dynamical effects and the progressive loss of spectral coherence as the origin of the spectroscopic signatures of H-bond symmetrization.

Efficient generation of entangled photons typically relies on spontaneous parametric down-conversion (SPDC) in phase-matched macroscopic nonlinear media. However, generating entanglement under phase-matching constraints requires additional bulk optics or interferometers. In contrast, ultrathin van der Waals semiconductors - such as transition metal dichalcogenides (TMDs) - exhibit strong enough optical nonlinearities for SPDC to be observed from subwavelength-thick media, thereby bypassing conventional phase-matching constraints. In this microscopic domain, the intrinsic crystal symmetry governs the nonlinear optical response, enabling the native generation of polarization-entangled photon pairs. However, generating these states efficiently has been fundamentally restricted by the material's coherence length ($L_c$), which limits the attainable conversion efficiency. Here, we investigate periodically-poled TMDs (PPTMDs) designed to scale up this interaction via quasi-phase matching. We demonstrate that mechanically flipping the sign of the nonlinearity at precise intervals of $L_c$ introduces quasi-phase matching, that scales the pair-production rate while preserving the pristine, symmetry-generated polarization entanglement, with fidelities exceeding 99%. Backed by a rigorous theoretical model, our work clarifies the interplay between crystal symmetry and propagation effects in thin nonlinear media, providing a new avenue for engineering quantum light in nanophotonic systems.

In this article, we connect statistical inference for spatial point processes with the analysis of waiting pedestrian crowds through two interconnected contributions. First, on the methodological side we develop an inference procedure for semiparametric spatial point process models leveraging replicated spatial patterns, i.e., multiple approximately independent realizations from the same process. Second, we show that spatial point processes provide a suitable modeling framework for waiting pedestrians, capturing two key aspects: spatial inhomogeneity driven by location attractiveness and repulsive interactions between pedestrians. These two components are central to the inference problem itself, since spatial point process modeling hinges on disentangling background intensity from interaction. Although replicated spatial patterns are rare in point process literature, they are available here through a unique real-life pedestrian dataset, thereby directly linking the methodological development to the physical application. We use the proposed methods to fit and evaluate determinantal and Gibbs point processes in a simulation study and a real-world case study. Despite persistent challenges in decoupling the influences of inhomogeneity from interaction, these models are able to reproduce key empirical features of waiting pedestrians.

We report multi-angle reflectivity measurements in the extreme-ultraviolet (XUV) range for mono- and bilayer MoS$_2$ on a Si$_3$N$_4$ substrate. Using a single-sheet 2D conductivity model, we extract the complex optical response of the MoS$_2$ bilayer between 25 and 90 eV and derive an effective refractive index by introducing a thickness equal to the interlayer spacing. The MoS$_2$ monolayer response is consistently reproduced either by halving the 2D conductivity or the effective thickness, indicating a robust scaling with layer number. The resulting optical constants display a broad resonance at the Mo N$_{2,3}$ edge with no signatures of sharp core-exciton features despite the reduced dimensionality. First-principles calculations reproduce the experimental results and show that local-field (Hartree) effects dominate the XUV response, while screened-exchange (SEX) contributions remain weak and mainly induce spectral shifts. Our analysis demonstrates that excitonic effects play a minor role in the XUV optical response of atomically thin MoS$_2$, highlighting key differences with respect to the visible and infrared regimes, and calling for a reassessment of the use of Mo-based transition metal dichalcogenides in attosecond spectroscopy and XUV excitonics.

Reciprocity between a transmitter and receiver is a foundational requirement in wireless communications. A few recent works have suggested that reciprocity is broken under reflection by reconfigurable intelligent surfaces (RIS) when the reflection phase becomes incident angle dependent. In this work, we rigorously show that these claims are based on the use of idealized reflection coefficients that ignore mutual coupling between heterogeneous unit cells, surface-truncation effects, and structural scattering contributions from the RIS. Full-wave electromagnetic simulations of transmit/receive antennas and a finite-size RIS implemented via a particular unit cell design are performed to quantitatively demonstrate that reciprocity holds even in the presence of incident-angle dependent reflection phases. To show this, we calculate two-port antenna scattering parameters and evaluate the electromagnetic reciprocity integral to support our claims.

Cryogenic electron microscopy images sample individual biomolecules from their conformational landscapes, offering a route to infer the distributions underlying molecular mechanisms. However, because images are indirect measurements, they limit which features of an underlying landscape are statistically identifiable. In ensemble reweighting, this problem appears as a choice of resolution: conformational space is discretized into representative structures whose population weights are inferred from images. Adding structures increases nominal resolution, but nearby conformations may generate overlapping image distributions and indistinguishable weights. Here, we develop an information-theoretic framework that selects representative conformations by maximizing mutual information between ensemble weights and images under a probabilistic forward model. Analytically, we show in a one-dimensional Gaussian model that measurement noise sets the optimal spacing. Applied to molecular conformations sampled from simulation, the framework constructs near-optimal ensembles that span heterogeneity while avoiding redundancy. Thus, the measurement process induces a maximally learnable coarse graining of conformation space.

Electrochemical doping offers dynamic control of the electronic properties of organic semiconductors, and it is the enabling feature of a range of technologies, including electrochemical transistors, energy-storage devices, light-emitting electrochemical cells (LECs), and bioelectronics. Electrochemical doping is commonly controlled by the selection of the constituents in the active material of the device or the applied voltage bias, but herein we report that the incorporation of Au nanoparticles (Au-NPs) at an electrode interface can constitute an alternative control parameter. The LEC features balanced p- and n-type electrochemical doping that forms a p-n junction doping structure in its active material, and we find that it is possible to reshape this doping profile by incorporating Au-NPs at an electrode interface. Specifically, we establish that the inclusion of neat non-capped Au-NPs at the anodic interface shifts the p-n junction (i.e., the emission zone) away from the anode. In contrast, the inclusion of Au-NPs capped with sodium citrate is found to reverse this behavior, so that the emission zone is instead moved towards the anode. We utilize this control parameter to shift the emission zone towards a position of constructive (destructive) interference, as manifested in a strong increase (decrease) of the LEC emission efficiency. Our findings establish an interfacial strategy for modulating the spatial profile of electrochemical doping and tuning device performance without altering the chemistry of the active material, relying instead on the surface modification of one electrode. This approach is important because it provides a versatile and minimally invasive route to optimize electrochemical devices while preserving the intrinsic properties and formulation of the active material.

Materials discovery increasingly relies on the coordinated use of theory, computation, experiment, data-driven methods, and emerging quantum technologies, yet the full potential of these tools is realised only when they operate within workflows that reflect the complexity of real systems. This perspective summarises current capabilities, limitations, and opportunities across these domains, drawing on contributions from academia, industry, and national laboratories to identify the scientific and structural requirements for more reliable and efficient discovery. Classical simulations provide broad coverage across design spaces, while experimental measurements reveal degradation, heterogeneity, and kinetic processes that determine performance under realistic conditions. Machine learning accelerates exploration when supported by well-curated datasets with clear provenance and uncertainty quantification, and quantum computing offers promising routes into correlated electronic behaviour when aligned with properties that influence engineering decisions. Collectively, these insights highlight the need for reproducible workflows, shared data standards, realistic benchmarks, and a research culture that prepares scientists to work across paradigms. By integrating these methodological and organisational elements, the community can move toward discovery processes that deliver robust predictions, support confident decision making, and shorten the path from conceptual design to deployable materials.

Photon-number-resolving (PNR) detectors are essential components of photonic quantum technologies, yet thus far, no practical metric exists to certify how many photons they can genuinely resolve in a single measurement. Here we introduce an operational framework for quantifying the capability of a PNR detector to distinguish between different numbers of photons, i.e. its genuine resolution. In turn, we develop a practical and scalable protocol for certifying the genuine resolution of a detector, which is based on coherent state probes. We apply the method to a 28-pixel photon-number-resolving superconducting nanowire single-photon detector (PNR-SNSPD) and certify genuine four-outcome resolution. Our work highlights the critical requirements in terms of detector efficiency towards achieving high genuine resolution. This approach provides an operational benchmark for PNR detectors and fills a crucial gap in the characterization of photonic quantum devices.

Chiroptical sensing is central to gain fundamental insight into electronic, vibrational and rotational degrees of freedom of chiral molecules, and is a cornerstone for nanomedicine and drug discovery platforms. Current chiral sensing technologies to assess the enantiomeric imbalance of chiral pharmaceutical compounds are sensitive to ml volumes but are time-consuming and cannot be integrated on a chip, thus creating a major bottleneck for drug discovery and nanomedicine. Here, we propose a novel chiroptical sensing approach based on optical rectification in a photonic micro-cavity filled by a drug solution with nl volume. We theoretically demonstrate that, upon optical excitation by intense pulsed laser light, such a nonlinear effect produces a chirally-sensitive nV voltage burst at the electrically-gated micro-cavity boundaries, with sign depending solely on the drug enantiomeric imbalance. Our results shed light on the potential of optical rectification as a robust platform for innovative lab-on-a-chip devices enabling chiral sensing with nl sensitivity.

Simulating high-enthalpy fractured geothermal reservoirs is challenging due to the complex coupled processes of non-isothermal, multiphase, multicomponent flow, strongly nonlinear thermodynamics, and the dominant role of fractures. These complexities are amplified by mineral scaling, such as halite precipitation, which can impair reservoir permeability and well productivity. To address this, we present a new compositional flow model based on a persistent set of primary variables (pressure, enthalpy, and overall salt mass fraction). The formulation naturally handles phase transitions without manual switching, enhancing numerical stability. The model integrates a discrete fracture-matrix approach and employs an efficient, robust correlation-based phase-behaviour linearisation of saltwater thermodynamics, replacing expensive on-the-fly phase separation calculations. It incorporates the Kozeny-Carman relation to dynamically model porosity and permeability reduction from halite precipitation. Implemented in the open-source PorePy framework, the model is verified through a 1D salt dissolution benchmark against the established closed-source simulator CSMP++, showing strong agreement across geothermal conditions involving transitions between single- and multi-phase regions. Application to a 2D halite-saturated fractured reservoir with injection and production demonstrates the model's capability to predict halite precipitation patterns and their impact on permeability damage and energy recovery. Numerical results further show the model's value in predicting operational challenges such as wellbore blockage and the role of fracture connectivity. The model thus provides an open-source numerical tool for analysing complex heat and mass transport with mineral scaling in high-enthalpy fractured geothermal systems.

We have demonstrated that compact deformation sensors with heights and widths of about 100 micrometers can be fabricated by two-photon polymerization, using commercially available optical ferrules with embedded 125 micrometers-diameter optical fibers as the basic platform and the commercial photopolymers OrmoComp and FemtoBond as the fabrication materials.

Tunable, microscopic, and energy-efficient solutions for radio-frequency (RF) signal manipulation in the GHz regime are a key technology for efficient communication and sensing applications. Spin waves offer micrometer wavelengths at GHz frequencies, combined with strong magnetic-field tunability, making them inherently well-suited for tunable analog signal processing. Here, we demonstrate a novel concept: a micron-scale tunable RF phase shifter based on the wavelength shift of propagating spin waves. High energy efficiency is achieved by using the stray field of a micromagnet on a piezoelectrically actuated MEMS cantilever to locally induce this shift. The device shows a phase shift of more than 360° at a center frequency of 6.1 GHz using a phase-shifting area of less than 0.02mm$^2$. By changing the magnetic bias field, its functionality is experimentally confirmed over a range of center frequencies from 3 GHz to 8.2 GHz, and simulations show its applicability up to 14 GHz. A system-level characterization of an embedded device version demonstrates the qualification of magnonic phase shifters for highly integrated RF systems.

Scientific discovery in computational physics can often be framed as the optimization of quantitatively evaluable objectives subject to physical constraints. While researchers excel at formulating such problems, they frequently devote substantial effort to iterative refinement of methods and solution strategies. To accelerate this process, we introduce PhyNex, an autonomous agent that systematically explores the solution space of scorable scientific tasks by coupling large language model (LLM)-guided search with domain-specific computational tools that enforce physical consistency. PhyNex operates via progressive local search, accumulates reusable knowledge from both successful and failed attempts, and produces interpretable exploration trajectories that reveal which algorithmic components drive performance improvements. We validate PhyNex on three representative and scientifically important problems: predicting frequency-dependent dielectric spectra of semiconductors from crystal structure, designing probabilistic-circuit heuristics for Max-Cut on graphs, and optimizing charging protocols for Dicke quantum batteries in the chaotic coupling regime. Across the three tasks, PhyNex autonomously identifies solutions that match or exceed state-of-the-art approaches designed by human scientists, yielding search-averaged improvements of up to 3.8\% in spectral similarity, up to 15.0\% in normalized mean cut for Max-Cut, and 5.9\% in ergotropy at the $80\mathrm{k}$ training checkpoint in open exploration. These findings demonstrate that LLM-based agents with structured, feedback-driven exploration can substantially accelerate the path from problem specification to effective implementation, suggesting a practical division of labor in which scientists define objectives and constraints while automated systems navigate the methodological search space.

The advent of photonic integrated circuits (PICs) will allow the replacement of the large aperture of an optical telescope by a dense array of small apertures combined interferometically. The light coming from aperture pairs can be combined by a PIC in order to extract interferogram characteristics known as complex visibilities, from which the observed object can then be reconstructed. In such a compact interferometric imager, the optical components dedicated to image formation in a regular telescope are no longer necessary. In particular, such a concept is relevant for space missions where weight and size are critical. In this communication, we study such an instrument concept, focusing on signal-to-noise considerations. We recall the design basis for the field and the spatial resolution, and we show that the spectral resolution must be no less than the field to resolution ratio. Then, we analyze the signal-to-noise ratio of this concept, assuming that each spatial frequency is recorded only once, and compare the signal-to-noise ratio with that of a monolithic telescope. We perform the comparison in Fourier space for an identical number of recorded photons. We show that the noise propagation of the interferometric imager is identical to that of a monolithic telescope that would have a flat Modulation Transfer Function with a level roughly given by the ratio of the small apertures' diameter to the maximum baseline. We conclude that the noise propagation in low and medium spatial frequencies is unfavorable for the interferometric imager.

Holography offers unique advantages for delivering perceptual realism while preserving compact form factors in VR/AR. Its perceptual quality, however, hinges on encoding rich wavefronts of photorealistic scenes into interference patterns and then incoherently multiplexing the resulting wave fields for perception. Existing CGH paradigms decouple radiance estimation from wave propagation by pre-rendering radiance on discretized scene sectors. This separation between radiometric and wave-optical computation inherently limits the range of focus cues and visual effects that can be faithfully reproduced, including depth- and view-continuity, and physically based material behaviors such as glossy or mirror-like reflection and refraction. We present a physically accurate yet computationally efficient wave optics rendering framework leveraging path tracing to encode full 3D visual cues into phase holograms. Specifically, we employ a Monte Carlo method to solve both the rendering equation and the Rayleigh--Sommerfeld integral simultaneously. Our algorithm is fully compatible with modern graphics techniques and can generate multiple time-multiplexed random holograms with minimal additional time cost via Path Reuse. By employing a fast approximation with an ambient radiance cache, we realize an order of magnitude convergence speed improvement. The resulting coherent wave fields that inherently encode comprehensive visual effects are converted into phase-only holograms under complex-amplitude supervision. Through extensive simulations and experimental validations on a spatial light modulator-based display prototype, we demonstrate faithful holographic reconstructions of natural 3D cues and complex materials, including realistic defocus blur, view-dependent effects, as well as appearance highlights and reflections.

Coarse-grained (CG) molecular simulations offer an efficient alternative to atomistic molecular dynamics to study large and complex biological systems. The accuracy of CG simulations has been increased dramatically by the introduction of machine-learned coarse-grained (MLCG) models. However, these models are typically designed to be used at a single thermodynamic point, lack temperature transferability, and can not be used to predict temperature dependent quantities like the heat capacity. Here we introduce a thermodynamically informed, temperature-transferable MLCG framework for proteins that explicitly decomposes the CG potential of mean force (PMF) into its energetic and entropic components. The model architecture enforces an exact thermodynamic relation between the energetic and entropic components of the PMF and guarantees physically consistent extrapolation and interpolation across temperature regimes. We validate this framework on an extensive dataset spanning a total of 250 $μ$s of molecular dynamics simulations across five temperatures between 300 K and 400 K for the Chignolin protein, and demonstrate that it reproduces the temperature dependency of the reference atomistic free energy surfaces, correcting the temperature-unaware baselines. Furthermore, we show that it is possible to apply an inexpensive, post-hoc temperature-dependent correction that does not require retraining the MLCG potential, accurately recovering the atomistic heat capacity at different temperatures. Overall, this work provides a physically grounded pathway toward thermodynamically transferable MLCG simulations of complex biomolecular systems.

Probing surfaces with optical beams is a powerful technique utilised for material characterisation, and more recently, measuring surface magnetisation with vortex beams has opened new avenues by utilising OAM beams for metrology. In this paper, we focus on extending the theoretical framework of unit order vortex beam reflection to spatially homogeneous magnetic surfaces, and derive expressions for the Goos-Hänchen (GH) and Imbert-Federov (IF) shifts. We calculate closed form expressions for the plane wave eigenpolarisations, polarisations which remain in the same vectorial direction after reflection, for both dielectric and homogeneous magnetic surfaces. These are found by extending the singularimetry formalism for dielectric surfaces, where the position of the phase vortex can be analytically approximated by taking into account the finite beam width, to homogeneous magnetic surfaces. From this we find the eigenpolarisations of the unit order vortex beam for reflection from both dielectric and homogeneous magnetic surfaces. We discuss with relation to the analytical forms of the Jones matrices for GH and IF shifts the additional reflection behaviour observed for homogeneous magnetic surfaces.

We study near-symmetric resonant stratified media to show how a tiny broken spatial symmetry can effectively be probed by structured light with or without orbital angular momentum. This is achieved by examining both the in-plane and out of plane Goos-Hänchen and Imbert Fedorov shifts, respectively, in the reflected light, magnified by resonant enhancement and weak value amplification. We show that non-reciprocity in reflection for illumination from opposite ends can result in different shifts, even to the extent of shifts with opposite signs for tiny imbalance resulting from the broken symmetry. We believe that our results can lead to new type of extra-sensitive sensors for any agent (eg. refractive index, displacement, etc.) that can break the symmetry.

In this work, we propose a Nonlinear Pluggable Optic (NLPO) transceiver that combines the low latency and low power consumption of Linear Pluggable Optics (LPO) with the range and robustness of digital signal processing (DSP)-based transceivers. The proposed NLPO uses an analog photonic Next-Generation Reservoir Computing (NGRC) architecture, constructed on a photonic integrated circuit (PIC), to compensate for electrical-domain distortions as well as optical-channel impairments from chromatic dispersion and Kerr nonlinearity. Focusing on a simulated 50 GBd PAM-4 link, we find that the NGRC-based NLPO not only extends the range of LPO, but actually outperforms DSP-based solutions as well. Our simulations reveal two key advantages compared to DSP-based Intensity Modulation/Direct Detection (IM/DD) links: (1) the NGRC can take advantage of the optical phase information without requiring a local oscillator and (2) the NGRC can optically sample the transmitted data well above the symbol rate without requiring high-bandwidth electronics. This work showcases the potential for photonic NGRCs to outperform state-of-the-art digital solutions in real-world applications and opens a path to low-latency, lower-power IM/DD links at ranges of 10s of km.

More than a century of physics is recorded in the American Physical Society (APS) archive, but the corpus cannot be analyzed as a single, time-resolved object because its subject metadata are fragmented across eras with no shared vocabulary. We close this gap by using a frontier large language model to retrospectively assign the modern Physics Subject Headings (PhySH) to the historical archive, yielding a unified subject map for every APS paper from 1893 to 2025. The resulting map not only reproduces century-scale disciplinary arcs but also resolves the fine-grained lifecycles of individual ideas, materials, techniques, and discoveries across a vocabulary of over 3,000 PhySH Concepts. The map turns a fragmented archive into a quantitative substrate for systematic search and for data-driven studies of how physics evolves.

We demonstrate direct single-shot retrieval of the full in-plane vector field of broadband terahertz waves using electro-optic sampling. Implemented with a circularly polarized optical probe and polarization-sensitive camera, the method overcomes the projection constraints of conventional electro-optic sampling and enables direct time-domain reconstruction of transverse vector fields without Fourier-domain inversion or sequential polarization analysis. We experimentally validate the approach through spatiotemporal imaging of linear, circular, azimuthal and radial terahertz fields generated by polarization-engineered silicon metamaterials. The measurements reveal ultrafast time-resolved vector-field dynamics inaccessible to conventional scalar detection, including helicoidal field rotation within a single optical cycle and the spatiotemporal evolution of structured polarization topologies. Beyond structured-field imaging, we demonstrate single-scan vector-field spectroscopy through measurements of quartz birefringence and broadband terahertz waveplate performance. More broadly, this work transforms electro-optic sampling from a projection-based measurement into a vector-resolved imaging and spectroscopy platform for structured electromagnetic fields.

This paper studies an inverse problem in nonlinear optical encryption. We examine chosen-plaintext attacks (CPA) on a nonlinear optical encryption strategy that integrates double random phase encryption (DRPE) into a nonlinear optical propagation model to enhance the security of the combined system. We first demonstrate that the system's phase information can be decoded from carefully designed differential CPA data. We then demonstrate that the strength of the optical device's nonlinearity can also be recovered from CPA data, indicating that including this parameter as an additional security key does not enhance protection against CPA attacks, although numerical simulations show that strong nonlinearity still poses significant challenges for CPA attacks. Finally, we provide a stability analysis to demonstrate that small errors in decoded security keys result in only small errors in the decrypted text, even though the encryption process is nonlinear.

The Fermi paradox is sharpened, not weakened, by plausible extrapolations of artificial intelligence, autonomous robotics, in-situ resource utilization, orbital manufacturing, space-based computing, and uncrewed interstellar probes. Once a civilization can design, launch, and maintain autonomous industrial systems beyond its home planet, interstellar expansion no longer requires biological starships or a human-like empire. It can proceed through low-mass probes, robotic seed factories, archival payloads, biological repositories, local computation, and slow replication across nearby stellar systems. This paper proposes the quiet expansion filter: old, stable civilizations that reached autonomous AI-cosmoindustry probably did not arise in the part of the Galaxy capable of reaching the Solar System, because after that threshold interstellar expansion becomes too useful, inexpensive, and rational for all civilizations to refuse; however, successful expansion would be machine-mediated, distributed, low-noise, and partly biological rather than Kardashev-like or imperial. Order-of-magnitude estimates indicate that a single post-threshold civilization could saturate its reachable stellar neighborhood within ~10^7 yr -- less than 0.1% of Galactic age -- at modest energy cost per probe. The novelty of the proposal lies not in any new mechanism but in extending the AI-filter literature toward post-threshold observability predictions. The hypothesis predicts that successful advanced expansion, if present, is more likely to appear as weak artifacts, local probes, small-scale resource processing, exoplanetary anomaly clusters, or techno-biological preservation systems than as galaxy-scale energy harvesting.

In this work, we derive the working equations for the transition density matrices within the equation-of-motion frozen-pair coupled-cluster framework. We focus specifically on the recently developed EOM-fpCCSD and EOM-ptCCSD models and present the corresponding transition dipole moments and oscillator strengths obtained via approximate expressions. This formulation avoids the computational overhead of solving the coupled-cluster $Λ$ equations. Furthermore, we employ a matrix inverse approximation to eliminate the need for explicit calculation of the left EOM eigenvectors. The accuracy of the resulting EOM-fpCCD and EOM-ptCCSD excited-state properties is benchmarked against the linear-response (LR)-CCSD method. Our results demonstrate that the description of excited state properties is improved when using EOM-fpCCSD and EOM-ptCCSD models compared to the standard EOM-CCSD variant.

Reducible oxides exhibit a rich interplay of electronic, structural, and chemical properties that underpins applications in catalysis, photovoltaics, batteries, and energy storage. This interplay is strongly shaped by excess electrons, often introduced by oxygen vacancies, that localize as small polarons and influence charge transport and surface chemistry. At surfaces, these polarons play a central role in charge localization, mobility, and reactivity, yet their finite-temperature dynamics remain difficult to access from first principles. Ab initio molecular dynamics is typically limited to picosecond time scales, precluding statistically meaningful sampling of polaron hopping dynamics. To overcome this limitation, we extend machine-learning-assisted polaron dynamics [V. Birschitzky et al., Phys. Rev. Lett. 134, 216301 (2025)] to redox-active oxide surfaces, using oxygen-deficient rutile TiO2(110) as a paradigmatic case. By accessing several nanoseconds of dynamics over a range of temperatures, we show that small-polaron mobility at the reduced rutile TiO2(110) surface is suppressed by several orders of magnitude relative to the corresponding bulk material, providing a microscopic interpretation of the lower electron mobilities observed in porous rutile TiO2 compared with single-crystal samples. This suppressed mobility arises from the loss of favorable hopping pathways: surface polaron motion is largely confined to planar inter-row trajectories within the second topmost layers, with only rare interlayer hopping events. Oxygen vacancies further reshape the polaron free-energy landscape by acting as attractive centers for excess electrons, biasing the polaron distribution toward nearby Ti sites and promoting occasional charge transfer to the outermost surface layer. These results establish a transferable machine-learning strategy for investigating polaron dynamics in reducible oxides.

We present a scalable Poisson solver with $O(N)$ complexity based on the fast multipole method (FMM) implemented in RAMSES. Our FMM constructs a hierarchy of FMM grids on top of the pre-existing Cartesian grid which is used to compute the force for hydrodynamics or particle-mesh simulations. In contrast to the $O(N)$ multigrid solver (MG) - an iterative method that requires multiple V-cycles through a multi-resolution hierarchy of Cartesian grids - the FMM algorithm performs just one upward pass through the same hierarchy, during which multipole expansions are accumulated and shifted, followed by a single downward pass, in which local expansions are propagated. Numerical tests indicate that FMM attains accuracy comparable to that of MG for smooth potentials and is particularly well-suited for problems with isolated boundary conditions, since it avoids the approximate Dirichlet boundary conditions required by MG schemes. Although in theory FMM requires around 30 times more floating-point operations than MG, its higher arithmetic intensity leads to comparable performance and better scalability relative to MG.

Efficient simulation of open quantum systems is central to modeling noisy quantum hardware and many-body dynamics. In trajectory-based tensor network methods, cost is often associated with trajectory-level quantities such as entanglement growth or bond dimension. However, the total cost of a fixed-accuracy simulation also depends on statistical sampling, and the interplay between per-trajectory complexity and sampling effort remains poorly understood. Here we introduce a cost-resolved framework for matrix product state (MPS)-based quantum trajectory simulations that decomposes total cost into memory per trajectory, runtime per trajectory, and sampling effort. We show that physically equivalent stochastic unravelings of the same Lindblad dynamics do not necessarily reduce total cost, but instead redistribute cost between trajectory complexity and statistical convergence. This trade-off is quantified by two dimensionless inflation factors: a bond dimension inflation $α$ and a sampling inflation $κ$, which together determine the preferred unraveling under hardware-dependent memory and parallelism constraints. We provide a practical protocol for extracting $(α,κ)$ from modest pilot simulations and demonstrate it using benchmarks across multiple noise channels. The resulting decision maps show that the computationally favorable unraveling can change with noise strength, time-step resolution, system size, and available parallelism. These results establish unraveling choice as a hardware-aware simulation design problem rather than an intrinsic optimization of trajectory entanglement alone.

We often decide how to treat friends based on observations of their past behavior, whereas actions toward strangers are typically guided by their public reputations. These two kinds of information underlie two classical mechanisms for the evolution of cooperation$\unicode{x2014}$direct and indirect reciprocity$\unicode{x2014}$which have largely been studied in isolation. They are not interchangeable: we can recall the past actions of only a small circle of close contacts, whereas for the far larger pool of strangers we must rely on public reputations. Here we develop a mathematical framework built on this distinction. Each individual engages in direct reciprocity in local games within a finite neighborhood of friends, whose actions they observe directly, and in indirect reciprocity in global games with a large population of strangers, known only by reputation. Separating local and global interactions allows us to address two questions. First, can cooperation persist under a cognitively simple norm of judgment? We show that combining direct and indirect reciprocity resolves the scoring dilemma: conditional cooperators resist invasion by both unconditional cooperators and unconditional defectors, where indirect reciprocity alone would fail. Second, how should one treat a friend whose past behavior conflicts with their public reputation? We find that the strategies that maximize cooperation are forgiving$\unicode{x2014}$overlooking whichever piece of information is unfavorable$\unicode{x2014}$and that these forgiving strategies can often remain robust to invasion. By distinguishing between local and global scales of interaction and integrating information across them, our framework offers a more cognitively realistic account of how reciprocity sustains cooperation.

A simple electrostatic two-equation model for dual-layer thin-film transistor (TFT) operation is developed. The model distributes electrostatic charge between the top and bottom semiconductor layers, and the resulting transfer and mobility curves accurately simulate experimental dual-layer a-IGZO/a-IZO TFT operation. The model further provides an analytic expression that maps charge confinement in the high-mobility a-IZO bottom semiconductor layer with the a-IGZO top-layer thickness and the conduction-band offset. By considering both a-IGZO/a-IZO layer charge partition and competing thickness-dependent oxygen vacancy trap density effects, the model suggests an optimal a-IGZO layer thickness of 9 to 12 nm. Importantly, this general electrostatic model extends to most dual-layer TFT systems and calculates how the top semiconductor layer TFT turn-on voltage changes sharply with the conduction band offset and layer thickness.

In radiation detectors, the spatial distribution of the electric field plays a fundamental role in their operation. Access to this field distribution is of strategic importance, especially when investigating the perturbing effects induced by incident radiation. For example, one dangerous effect that prevents their proper operation is the accumulation of internal space charge. Here, we probe the two-dimensional electric field in a Schottky CdTe detector using the Pockels effect and report on its local perturbation after exposure to an optical beam at the anode electrode. Our electro-optical imaging setup, together with a custom processing routine, allows the extraction of the electric-field vector maps and their dynamics during a voltage bias-optical exposure sequence. The results are in agreement with numerical simulations, allowing us to confirm a two-level model based on a dominant deep level. Such a simple model is indeed able to fully account for both the temporal and spatial dynamics of the perturbed electric field. This approach thus allows a deeper understanding of the main mechanisms affecting the non-equilibrium electric-field distribution in CdTe Schottky detectors, such as those leading to polarization. In the future, it could also be used to predict and improve the performance of planar or electrode-segmented detectors.

Planar turbulent mixing layers, formed by the interactions of two parallel streams with different velocities, have been studied far more than three dimensional (3D) turbulent mixing layers, in which the incoming streams are skewed, and not parallel. Yet many practical shear flows are 3D. Here, we develop and validate an experimental methodology to generate and characterize skewed turbulent mixing layers and to quantify how mean-flow skew modifies mixing layer dynamics. We introduce skew with a spanwise deflection of the mean flow using turning vanes mounted near the trailing edge of a splitter plate, and we use cross-wire anemometry to investigate the downstream evolution of the flow. Relative to the planar configuration, the skewed mixing layer exhibits systematic reductions in both mean and turbulent quantities, with deviations reaching approximately 40\%. Despite these quantitative differences, the fundamental characteristics of the mixing layer remain largely unchanged. Mean-velocity profiles collapse under similarity scaling, shear-layer thicknesses retain approximately linear downstream growth, and Reynolds-stress profiles preserve their characteristic near-Gaussian form. Townsend's structure parameter, which quantifies the efficiency of turbulent momentum transport, remains approximately invariant between the planar and skewed configurations, in contrast to skewed turbulent boundary layers, wherein comparable mean flow skewing reduces the parameter by approximately 30\%. These results indicate that mean flow skew modifies turbulent mixing layers quantitatively while exerting only a secondary influence on their underlying dynamics. This study establishes a controlled experimental framework and empirical benchmark for future investigations of three-dimensional free-shear turbulence.

Space-time metamaterials that exhibit synthetic motion promise arbitrary control over the momentum, frequency and energy of scattered light, but realising the required space-time modulation in a programmable way remains a challenge. Here we program synthetic motion using a single spatial light modulator in a 4f geometry which imprints a continuously tunable pulse-front tilt onto a high-intensity pump pulse, inducing reflectivity modulations at a sub-wavelength indium tin oxide thin film with synthetic velocities spanning the sub- and superluminal regimes. The angle-resolved spectrum of a scattered probe pulse reveals space-time diffraction patterns whose gradient and bandwidth vary continuously with synthetic velocity, in excellent agreement with theory. Splitting the shaped pump into two independently controlled pulses yields space-time double-slit diffraction with tunable fringe separation and frequency-momentum gradient. This programmable platform opens a path towards non-linear and periodic space-time trajectories for tabletop analogue studies of relativistic phenomena and space-time metasurfaces.

Two-dimensional magnets offer compelling platforms for spintronics and quantum technologies, yet predicting their magnetic ground states, moments, and anisotropy remains challenging. This limitation primarily arises because existing machine-learning representations encode chemical environments without capturing the symmetry or exchange physics that govern magnetism. In this work, we introduce the symmetry-electronic fingerprint (SEF), a physically interpretable representation that encodes crystallographic symmetry operations, Wyckoff-site geometry, together with site-resolved electronic structure. Combined with ensemble learning with random forests, the SEF accurately classifies magnetic ordering while regressing moments alongside anisotropy energies while simultaneously resolving the distinct regimes of itinerant Stoner ferromagnetism from localized superexchange. What sets the SEF-trained models apart is that regions of elevated model uncertainty are not a failure but a diagnostic, identifying materials where these mechanisms compete. First-principles calculations on Co- and Ni-based halides and oxides confirm that these regions correspond to genuine near-degenerate FM and AFM phases with magnetic frustration, suppressed anisotropy, and emergent non-collinear ordering. By encoding symmetry together with exchange physics directly into the representation unlike conventional descriptors, the SEF transforms model uncertainty into a compass pointing toward two-dimensional materials where small perturbations drive transitions between collinear, frustrated, or non-collinear magnetic phases.

We theoretically investigate the transmission and diffraction of a weak probe field from an electromagnetically induced grating formed in a weakly driven two-level medium coupled to engineered quantum reservoirs. Using a perturbative solution of the optical Bloch equations in the weak-driving regime, we analyze how normal-vacuum, thermal, and broadband squeezed-vacuum environments modify the probe susceptibility and consequently reshape both the spatial transmission function and the far-field diffraction patterns. We show that reservoir statistics have a pronounced impact on the diffraction response by altering the amplitude and phase of the induced grating. Thermal reservoirs enhance the transmission modulation and increase the intensity of the dominant diffraction orders, whereas squeezed-vacuum reservoirs generate strongly phase-sensitive modifications that selectively redistribute optical power among diffraction channels. We further demonstrate that the detuning between the squeezed reservoir and the driving field provides an efficient mechanism for controlling diffraction directionality, leading to substantial amplification of selected angular orders. In two-dimensional geometries, squeezed-vacuum correlations produce highly structured phase landscapes and strongly anisotropic diffraction patterns, enabling directional enhancement of specific diffraction channels while suppressing others. These results establish reservoir engineering as a versatile approach for controlling transmission, diffraction efficiency, and angular selectivity in minimal two-level systems, with potential applications in programmable photonic devices, beam steering, and quantum optical platforms.

Omnidirectional flat-band resonances, characterized by an enhanced photonic density of states and inherent angular robustness, are highly sought-after in integrated nanophotonic devices, particularly when integrated with chiral functionality. Here we realize such resonances in a nonlocal silicon membrane metasurface patterned with periodic square-lattice air-hole arrays. Increasing the lattice period not only compresses the Brillouin zone but, crucially, weakens the evanescent coupling between neighbouring Bloch modes associated with the same-order guided resonances. Driven by the tight-binding model in the limit of weak inter-unit-cell coupling, the pronounced band flattening of the degenerate guided resonance along both $k_{x}$ and $k_{y}$ yields, giving rise to an omnidirectional flat-band resonance. Remarkably, both numerical simulations and experiments reveal a universal route for endowing flat-band guided resonances with optical chirality through the deliberate breaking of the mirror symmetry of air holes. As a result, the omnidirectional chiral flat-band resonance emerges along both principal in-plane directions, with $Q$-factors exceeding 10$^{3}$ and circular dichroism greater than 0.9 over a wide angular range of $\pm 5^{\circ}$. Nonlinear measurements further show that the resulting resonance not only drives highly efficient third-harmonic generation but also imparts a pronounced spin-selective character to the nonlinear process. Simultaneously, the highly efficient nonlinear process also enables chirality-controlled frequency-upconversion imaging. Our results establish a general paradigm for engineering omnidirectional chiral flat-band resonances in planar silicon platforms, opening new opportunities for nonlinear nanophotonics and chiral imaging.

Picture naming requires the transformation of visual object information into a spoken lexical response through perceptual, semantic, lexical, and articulatory processes. This study asked whether semantic-category information is recoverable from high-density EEG during overt picture naming. Sixteen native French-speaking participants performed a picture-naming task using line drawings. Picture labels were embedded with a multilingual text-embedding model and organized into nine interpretable semantic categories, providing a data-driven semantic target space for neural decoding. EEG activity was represented channel-wise using a pre-trained single-channel EEG encoder over an early post-stimulus window, a later naming-related window, and their combination. Nine-class decoding showed above-chance semantic-category discrimination in all temporal representations. Balanced accuracy increased from 0.562 in the early window to 0.610 in the naming-related window, and reached 0.781 when both windows were combined, with a maximum Macro-F1 of 0.784. Class-level F1 scores showed consistent gains across semantic categories, and sensor-level decoding maps indicated spatially distributed category information. These findings suggest that semantic-category structure is reflected in EEG activity during overt picture naming and that early and naming-related temporal windows provide complementary information. The results support the use of modern neural decoding methods as tools for investigating lexical-semantic processing in spoken language production.

Agent-based modelling is gaining recognition as a powerful approach for simulating complex cellular pathways, owing to its ability to reproduce emergent biological behaviours without requiring extensive kinetic parameterisation. In this article, we present a GPU-accelerated agent-based simulator specifically designed to model and analyse signalling pathways involved in cancer progression, and to evaluate therapeutic interventions. Our approach leverages the computing capabilities of FLAME GPU 2, a GPU-accelerated agent-based modelling framework, to efficiently manage simulations involving millions of molecules interacting within a three-dimensional environment. Each molecule is represented as an autonomous agent with defined physical properties, capable of binding, releasing reaction products, migrating between compartments, and interacting based on spatial proximity. An intuitive graphical interface supports model construction, parameter setup, and real-time modification of treatment strategies. As the primary focus of this paper, we validate the simulator on the MAPK/ERK cascade affected by the BRAFV600E mutation, demonstrating that it accurately reproduces dose-response trends observed in clinical data and outperforms both deterministic models and our prior agent-based implementations. A second case study extends the approach to nuclear signalling by reproducing the dynamics of cFos expression and phosphorylation. This demonstrates the simulator's ability to capture compartmentalised regulation, reproducing transient mRNA responses and protein accumulation, including the effect of an unresolved negative transcriptional regulator. Together, these results show that GPU-accelerated ABM can faithfully replicate both drug response and emergent gene expression dynamics, providing a scalable and biologically grounded computational tool for supporting precision oncology.

In this note, we show that, for all $n\ge 2$, the number of distinct rooted binary phylogenetic $X$-trees displayed by a binary tree-child network $\mathcal{N}$ on $X$ with $n$ leaves is at most $2^{n-1}-1$ and that this upper bound is sharp. Furthermore, if $\mathcal{N}$ displays exactly $2^{n-1}-1$ such trees, then exactly one rooted binary phylogenetic $X$-tree is displayed twice, and this tree can be canonically found by iteratively replacing a reticulated cherry with a cherry.

The capacity of dynamical systems to resist and recover from perturbations, broadly referred to as resilience, is commonly expressed by two complementary quantities: ecological and engineering resilience. As many complex systems exhibit critical transitions, or tipping points, understanding how these resiliences jointly change nearby them is central to characterising and anticipating such shifts. Here, we develop a theoretical framework that clarifies this for bifurcation-induced tipping, i.e., critical transitions triggered by the crossing of a local bifurcation. Using normal form theory, we derive explicit scaling laws for commonly used resilience metrics as functions of the distance to the bifurcation point in parameter space, and show that these extend to general models up to a scaling factor. They are particularly relevant for detecting tipping, where the relative behaviour of metrics matters more than their absolute values. The rates at which metrics decrease as the bifurcation is approached depend on both the type of bifurcation and the metric considered. Furthermore, our results show that, sufficiently close to a local bifurcation, resiliences are intrinsically linked. Our predictions, which replace previously proposed scalings based on heuristic arguments, are validated for three representative models covering all commonly encountered local bifurcations in one-dimensional systems.

Lipid metabolism is a central biological process that is commonly studied using destructive mass-spectrometry experiments. A recently proposed strategy, uses multiple labels to extract temporal information about lipid metabolism from a single destructive measurement. However, the computational complexity of the model-based data analysis increases rapidly with the number of labels, creating a fundamental trade-off between the information content of the measurements and the cost of analysis. Here, we examine how the number of modelled labels affects parameter estimation accuracy, trajectory recovery, and computational cost, and whether modelling fewer labels than are experimentally available can mitigate this trade-off. Using synthetic data from a five-label experiment, we find that modelling three of the five labels provides a practical balance between experimental feasibility, inferential power, and computational tractability. In an application to hepatocyte triglyceride cycling, we further show that the most cost-efficient, single-label model can yield biologically implausible predictions for unobserved species, whereas models that resolve more labels better constrain these latent dynamics. These results provide practical guidance for selecting model resolution in multi-label experiments and establish a quantitative basis for balancing inferential power against computational cost.

A microwave linear analog computer (MiLAC) is a tunable microwave network that performs computation through wave propagation in the analog domain. In beamforming, data streams pass through a reconfigurable admittance network and emerge as antenna signals. For communications, MiLACs are preferably lossless and reciprocal to avoid power dissipation and non-reciprocal components, but these constraints limit the analog beamformers they can realize. Fully-connected MiLACs offer broad flexibility at the cost of a quadratic number of tunable admittances in the antenna count. Stem-connected MiLACs reduce this scaling to linear and preserve point-to-point capacity, but their role in multiuser downlink beamforming and under bounded, discrete hardware constraints has remained open. This paper addresses both questions for the multiuser multiple-input single-output downlink. We show that a stem-connected MiLAC can realize every beamformer on the complex Stiefel manifold and prove that, when $N\ge 2K-1$, this Stiefel-restricted design achieves the same sum-rate as the fully-connected MiLAC, where $N$ and $K$ are the numbers of transmit antennas and users. We then develop a weighted minimum mean-square error solver with a Riemannian Stiefel update, together with a closed-form projection baseline and an alternating refinement for bounded, discrete susceptances. Simulations show that the stem-connected MiLAC matches fully-connected MiLAC performance, approaches the fully digital sum-rate upper bound without symbol-rate digital processing, and recovers most of the loss caused by direct hardware-grid quantization.

Cooperative massive multiple-input multiple-output (MIMO) promises large gains over cellular deployments, but existing comparisons of different architectures often mix antenna distribution, inter-site coordination, and processing assumptions. This paper introduces a graph-based framework for fair comparison of cellular, coordinated, and cell-free massive-MIMO systems. We differentiate between two key properties, namely antenna distribution and inter-site cooperation, which yields seven representative system types. We derive compatible uplink and downlink spectral efficiency (SE) expressions, including an uplink bound for detectors with mixed instantaneous and statistical effective channel state information (CSI), and adapt scalable user association and processing rules to all considered architectures. We evaluate these systems using extensive numerical simulations and show that for a fair comparison much larger simulation areas (at least 2.5 $\times$ 2.5 km2) than commonly used are required. We introduce the relative capacity, which measures how closely each architecture approaches centralized cell-free processing. The results show that coordinated, phase-aligned beamforming across spatially distributed antennas is the main source of cooperation gains. In dense deployments with few antennas per access point (AP), coordinated Distributed Antenna System (DAS) and hybrid cell-free architectures achieve much of the centralized cell-free performance while requiring substantially weaker midhaul assumptions.

Reliable return navigation remains an important challenge in GPS-denied environments where external positioning infrastructure may be unavailable or unreliable. This paper presents ForestBack, an infrastructure-free pedestrian return navigation framework based on breadcrumb-based pedestrian dead reckoning (PDR). The system records a user's walking route as a sequence of reversible breadcrumb nodes and generates reverse-path guidance without requiring GPS, Wi-Fi, Bluetooth beacons, or pre-installed infrastructure. ForestBack integrates acceleration-based step detection, adaptive step-length estimation, magnetometer-assisted heading estimation, barometric-altitude correction, and bidirectional breadcrumb path reconstruction. The system was evaluated using an indoor obstacle-avoidance route with five checkpoints, where the user navigated around a central obstacle. A dataset of 36 walking trials and 42,474 time-series samples was used for evaluation, including IMU signals, magnetometer readings, barometric variables, turn-event labels, ground-truth trajectories, baseline PDR outputs, proposed ForestBack outputs, and power-related measurements. Experimental results show that ForestBack reduced the mean RMSE from 1.129 m to 0.965 m compared with traditional PDR, corresponding to a 15.76% improvement. The mean final-position error was reduced from 1.781 m to 1.388 m, while turn-event detection consistency reached approximately 99.90%. These results indicate that ForestBack improves trajectory reconstruction and route-preserving return guidance in obstacle-avoidance scenarios. The released dataset and analysis notebook support reproducibility and future benchmarking of infrastructure-free PDR-based return navigation systems.

This paper investigates the impact of reciprocity calibration errors on the downlink spectral efficiency (SE) of multi-user large antenna systems. Specifically, we consider two calibration approaches: (a) global calibration, in which all antennas (can be distributed access-points (APs)) in the system cooperatively perform calibration, and (b) local calibration, wherein only a subset of antennas involved in downlink beamforming performs calibration. We derive the downlink SE considering the use-and-then-forget bound and side-information bound, and then demonstrate that, when downlink pilots are employed (in the case of side-information bound), the global calibration outperforms local calibration for arbitrary calibration topologies.

Reciprocity-based downlink beamforming is imperative for a scalable time-division duplex massive multiple-input multiple-output~(MIMO) deployment. Specifically, for a dual-antenna repeater-assisted massive MIMO system, a mismatch between forward and reverse path gains at the repeater can exacerbate the overall calibration error between the user equipments (UEs) and the base station (BS), which potentially also contains calibration errors of their individual radio-frequency chains. This paper models the effects of such calibration errors, underpins the relations between the uplink and downlink channels for repeater-assisted systems with calibration errors clubbed with the over-the-air channel estimation errors, and derives analytical expressions of the downlink spectral efficiency. The presented results can then be simplified to several special cases, underscoring situations wherein such errors can become pronounced.

Effective modeling of endangered marine mammal species, such as the North Atlantic Right Whale, is critical for balancing marine conservation with the growing blue economy. Passive acoustic monitoring data collected by autonomous underwater vehicles provide new opportunities for localized marine species detection and oceanographic sensing, but introduce complex statistical challenges such as zero inflation, imperfect detection, and intricate dependence structures. In response, we propose the Deep Zero-Inflated Bernoulli (DeepZIB) model--a deep statistical method which jointly models latent species presence and conditional detection probabilities while learning complex habitat relationships from heterogeneous covariate information. We establish theoretical results on the model's structural properties and conduct simulation experiments to demonstrate its ability to recover underlying parameters and latent presence fields. Application to real-world passive acoustic monitoring data on the North Atlantic Right Whale along the U.S. East Coast demonstrates improved model adequacy and predictive performance in capturing the species' dynamic and spatially varying habitat. A key advantage of DeepZIB is its ability to generate high-resolution, spatially and temporally varying presence maps, providing valuable insights for targeted and risk-aware management of blue economy industries, ranging from offshore and marine energy, to fisheries management and maritime transport.

The Geometric Dilution of Precision (GDoP) characterizes the noise sensitivity of a Time-Difference-of-Arrival (TDoA) localization system, but does not capture every way the analytical multilateration solution can become ill-conditioned. We introduce a complementary geometry-quality metric $κ$, the leading coefficient of the closed-form TDoA solver's quadratic, and derive its $N$-dimensional generalization through a vectorized formulation. Two closed-form algebraic identities relate $κ$ to the Jacobian determinant of the measurement model and to the quadratic's discriminant, establishing that the system exhibits exactly two distinct singularity loci: branch divergence and the Jacobian/branch-merge locus flagged by GDoP. A Cramér--Rao-bound-linked closed form for the noise sensitivity $σ_κ$ under the standard Gaussian ToA model is validated against Monte Carlo to 2% median relative error. An empirical atlas over a dimensionless geometry parameter space confirms both identities at machine precision and shows that $κ$-bad regions and GDoP-bad regions are non-trivially disjoint in target space, establishing the two metrics as genuinely complementary. A case study on a four-node operational array, with per-sensor time of arrival (ToA) noise estimated empirically from Automatic Dependent Surveillance Broadcast (ADS-B)-paired over-the-air captures, shows that the theory-predicted threshold and a Monte-Carlo-measured operational threshold agree on the per-subsystem ordering at the deployment noise level. Their ratio is approximately constant across the three two-dimensional subsystems, serving as a deployment-specific calibration constant between the algebraic $κ$-noise floor and the downstream operational threshold, analogous in spirit to the standard relation linking GDoP to the circular error probable.

In recent years, neural networks have become the dominant models in most point cloud upsampling methods. Although these approaches are achieving good results, they do have drawbacks, such as a lack of interpretability and data dependency. Moreover, they have to be trained on a dataset that is similar to the test data in order to perform well. To avoid these disadvantages, we propose Point Cloud Upsampling through Patch-based Frequency Superposition (PUtPFS), an optimization-based approach that selects subsets of points and estimates the surface of this set through superpositioning spatial frequencies. Then, new points are placed on this surface. By successively selecting points in the least dense regions of the point cloud, a uniform upsampling can be reached. With this method, we surpass the current best upsampling results in the commonly considered point-to-surface distance. Furthermore, we achieve the best Chamfer and Hausdorff distance among the optimization-based approaches. As an additional advantage, our method does not need any training data and is mathematically interpretable.

Integrated Sensing and Communication (ISAC) enables sensing capabilities by reusing communication signals, making it particularly attractive for large-scale deployments through signals of opportunity. While most existing ISAC research targets wideband systems, Low Power Wide Area Network (LPWAN) technologies such as LoRa remain largely unexplored from a radar-like sensing perspective. Existing LoRa-based approaches mainly focus on motion detection or require modifications of the communication waveform, limiting their applicability in deployed networks. This paper investigates the feasibility of radar-like sensing using unmodified LoRa communication signals as signals of opportunity in a purely passive bistatic ISAC configuration. The proposed approach focuses on Doppler-based sensing to enable target separation and super-resolved target estimation without interfering with existing LoRa network operation. The analytically derived sensing capabilities are compared against simulation results and validated through bistatic measurements using two USRP B210 software-defined radios, confirming the feasibility of Doppler-based LoRa sensing under practical conditions and revealing relevant implementation challenges. The results demonstrate that LoRa-based ISAC enables highly scalable, large-area, low-resolution sensing by leveraging existing infrastructure, providing a complementary sensing capability to area-limited high-resolution 6G ISAC systems, and a foundation for future multi-node and data fusion extensions.

The intelligent evolution of mission-critical networks, such as the Internet of vehicles (IoV) and the low-altitude economy (LAE), requires sixth-generation (6G) networks to move beyond discrete physical parameter estimation toward deeper environmental understanding. However, existing integrated sensing and communications (ISAC) studies mainly focus on target-level sensing, which provides fragmented snapshots of the physical world and lacks the behavioral semantic capability to interpret intent. This limitation hinders the intelligent evolution of such networks and prevents 6G from acquiring the essential sensing foundation to evolve into an "intelligent service engine". To bridge this gap, ISAC must advance toward event-level sensing, which models continuous-time states to enable persistent recognition and prediction of target intent and behavioral semantics. This article presents a comprehensive overview of event-level sensing in 6G ISAC networks. We first introduce its fundamental concepts, sensing types, and representative scenarios. We then review key enabling techniques across waveform design, target state estimation and tracking, and event recognition. Furthermore, focusing on IoV and LAE scenarios, we discuss representative applications of ISAC event-level sensing and the intelligent enhancement of downstream operational functions enabled by event-level information. Finally, we highlight future research trends and potential directions to further advance ISAC event-level sensing toward intelligent and proactive 6G networks.

The rapid development of the low-altitude economy (LAE) has created growing demand for reliable aerial communication systems. Extremely large-scale multiple-input multiple-output (XL-MIMO) is a promising enabler for such systems due to its high spatial resolution and robust connectivity. However, three-dimensional (3D) mobility together with near-field propagation makes it difficult to obtain dedicated high-fidelity wireless datasets, hindering systematic algorithm development and evaluation. To address this issue, we develop LAETwin-XL, a digital twin (DT)-based toolchain and dataset for XL-MIMO research in LAE scenarios. Built on the Sionna ray-tracing (RT) module, the proposed toolchain simulates near-field and far-field channels with diverse wireless labels for practical environments. Building on this dataset, we further develop a conditional denoising diffusion implicit model (CDDIM)-based generative foundation model that is pretrained to learn transferable XL-MIMO channel representations from incomplete channel observations. Unlike conventional task-specific or foundation models that rely on relatively complete channel inputs, the proposed model can generatively infer informative channel representations from partially observed channels. Experimental results demonstrate that the proposed framework achieves effective zero-shot channel extrapolation performance. Furthermore, using lightweight task heads and limited training data, it enables parameter-efficient transfer to various downstream tasks (e.g., channel estimation, classification, and localization), delivering high accuracy and robustness even under sparse antenna observations. The codes and dataset are available at https://github.com/Lmyxxn/LAETwin-XL.

Graph Signal Processing (GSP) and Graph Neural Networks (GNNs) rely fundamentally on the matrix representation of the underlying graph topology. This representation defines key operators such as the graph Fourier transform, spectral filtering, and convolution. Existing parameterized operator families interpolate only partial subsets of classical graph matrices, while broader formulations become non-compact when representing transition-type operators, limiting both theoretical analysis and stable learning. To address this issue, we propose the Generalized Linear Graph Representation (GLGR), denoted by $\mathbf{Q}_{α,l}$, as a compact two-parameter operator family defined on a bounded linear domain. GLGR unifies major classical operators together with transition-type operators without requiring asymptotic parameters. Theoretically, we show that $\mathbf{Q}_{α,l}$ admits a variational decomposition balancing local smoothness and global degree-weighted energy, derive spectral perturbation bounds, and establish graph-aware sufficient conditions for positive semi-definiteness. Building on this formulation, we develop Adaptive GLGR Convolution (AG-Conv), which makes the propagation operator itself learnable within end-to-end GNNs. Experiments on graph classification and node classification benchmarks show that GLGR improves both fixed-operator representation search and adaptive graph learning across multiple backbones.

Accurate joint tracking of mobile users, surrounding scatterers, and dynamic channels is a critical task for sixth-generation (6G) wireless systems, essential for both ensuring high-quality communications and empowering advanced selsing applications such as autonomous driving and immersive extended reality. While extremely large-scale multiple-input multiple-output (XL-MIMO) inherently offers strong support for this task through its high spatial resolution and spectral efficiency, its massive scale of antenna arrays, coupled with near-field propagation characteristics, makes joint trajectory and channel tracking time-consuming and hardware-intensive. To address these challenges, we rethink the problem from a vision-based signal perspective. Specifically, we design a subarray-based partially connected hybrid beamforming (PC-HBF) architecture with a tailored time-multiplexed (TM) mechanism. This effectively compensates for the aperture loss caused by limited radio frequency (RF) chains, generating high-fidelity Cartesian-domain signal images that inherently capture near-field spatial features. Based on this visual representation, we propose an improved CenterNet to perform accurate one-shot path localization, circumventing the path-iterative search required by conventional compressed-sensing-based methods. Building upon this to further improve the accuracy and exploit temporal correlation, a local small-scale orthogonal matching pursuit (OMP) refiner and a lightweight cascaded OMP tracker are developed. Finally, a Hungarian-based trajectory association module is incorporated to maintain track continuity and provide trajectory-level information for environment monitoring. Simulation results show that the proposed framework consistently outperforms representative baselines in position and channel tracking accuracy, especially under low-SNR and limited-hardware conditions.

Space-based GNSS reflectometry (GNSS-R) can detect terrestrial radio frequency interference (RFI) through elevated noise power in delay-Doppler map forbidden zones. This study evaluates how constellation size affects detection performance using Level 1 delay-Doppler observations from seven CYGNSS spacecraft collected over three months from the NASA PO.DAAC archive. Four metrics are analysed: detection latency, spatial coverage, spatial coherence, and persistence monitoring reliability. Results show that the full seven-satellite constellation reduces median detection latency by a factor of 4.7 compared with a single satellite and increases interception probability for a 5-minute emission from 2\% to 11.5\%. Median footprint revisit time improves from 5.8 hours to under 2.0 hours. Spatial coherence analysis indicates that a single satellite leaves up to 72\% of source structure unresolved. Persistence monitoring confirms interference onset 39 days earlier than single-satellite deployment. The largest gains occur between one and three satellites, establishing three satellites as the minimum effective constellation size.

High-accuracy positioning is critical for emerging applications such as autonomous driving, industrial automation, augmented reality, and smart cities. 3GPP Release 18 introduced carrier-phase (CP) positioning for 5G that offers superior accuracy compared to conventional time-based methods such as time of arrival (ToA). However, CP-based positioning requires resolving the integer phase ambiguity, which refers to the unknown number of full-wavelength cycles completed during signal propagation. Joint processing of ToA and CP can mitigate this integer ambiguity by narrowing down the search space of possible integers, particularly for short wavelengths. This paper investigates the performance of a positioning method that integrates ToA and CP measurements. As a main contribution, the analysis explicitly accounts for the error correlation between ToA and CP measurements. Furthermore, the study analyzes the impact of key 5G system parameters on positioning accuracy using this correlation-aware joint method in both factory and urban environments, where many 5G positioning applications are expected to emerge. The results highlight that exploiting this correlation can further improve positioning performance by approximately 7 percent. Moreover, the findings of this study provide insight into how 5G system parameters can be tuned to achieve centimeter-level accuracy under favorable conditions.

Learning graph topology of complex networks is challenging due to limited data availability and imprecise data models. Different from prior works that focus on structural priors with explicit control on macroscopic properties such as sparsity, this paper proposes a novel functional prior approach for graph topology learning. We postulate that complex networks are inherently optimized to perform a certain task (e.g., social networks specialize at optimizing a welfare function, biological networks are resilient towards node/edge deletion), which can be incorporated as a regularizer to assist in graph learning. Mathematically, we formulate a bilevel optimization problem where the lower-level problem solves the associated task on a candidate graph topology and the upper-level problem trades off between data fitting and task performance. We design a two-timescale gradient descent (TTGD) algorithm and show that under verifiable conditions, it finds a stationary point to the bilevel graph learning problem with a sublinear convergence rate. We provide theoretical insights on the graph topology learned from the functional priors and show that the resulting regularizers subsume a broad class of graph filter regularizers, including polynomial graph regularizers as special cases. We show via extensive experiments on synthetic and real datasets that the proposed formulation gives rise to reliable estimates of graph topology, even with insufficient data.

The U.S. has adopted four power regimes for opera tion in the shared unlicensed 6 GHz band -- standard power (SP), low-power indoor (LPI), geofenced variable power (GVP), and very low power (VLP) -- with maximum permitted EIRP levels of 36 dBm, 30 dBm, 24 dBm, and 14 dBm, respectively. Although these regimes are primarily intended to protect incumbent services, their heterogeneous transmit power levels also introduce additional coexistence challenges within 6 GHz Wi-Fi networks. In this paper, we develop an ns-3 Wi-Fi 6E/802.11ax coexistence testbed to study coexistence under heterogeneous power regimes and to provide a reproducible simulation methodology. To the best of our knowledge, prior work has not specifically examined self-coexistence issues within 6 GHz Wi-Fi networks. We evaluate two coexistence scenarios: one in which both the LPI AP and the SP AP are indoors, and another in which the LPI AP is indoors while the SP AP is outdoors. Results are compared against an indoor LPI--LPI baseline when applicable. Our findings show that: (i) the presence of an indoor SP AP can significantly degrade the goodput of an LPI AP; (ii) channel bandwidth is a key factor in determining the extent of SP-to-LPI impact, with the degradation being most severe at 20 MHz and partially alleviated at 160 MHz; (iii) physical blockage between outdoor SP and LPI APs improves fairness; and (iv) BSS coloring does not necessarily improve fairness in mixed-regime deployments. The simulation framework can be extended to study coexistence between Wi-Fi and cellular systems, as recently proposed by Ofcom in the U.K.

Standing on the shoulders of giants, we resolve the quantum pyramids conjecture, confirming the globally information-optimal measurement for an ensemble of equiangular equiprobable pure states, as conjectured by Englert and Řeháček (arXiv:0905.0510). We do so by proving the remaining entropy inequalities of Holevo and Utkin (arXiv:2506.06700), which certify optimality for obtuse and flat pyramids. For obtuse pyramids, our key contribution is a rigorous proof that local minimizers of the corresponding entropy inequality cannot have three distinct coordinate values. We show that eliminating this family can be reduced to a neat algebraic reciprocal inequality relating branches of the Lambert $W$ function, which may be of independent interest. For flat pyramids, we prove a tight $\ell^p$ inequality for zero-sum vectors that was recently conjectured, proved analytically in dimension $d=3$, and computationally verified for $d\leq 200$ by Holevo and Utkin (arXiv:2603.24017). We prove this bound for all $d\geq 2$ via a technique in symmetric inequalities known as the equal variables method.

Hole spin qubits provide one of the leading platforms for spin-based quantum computing due to their large intrinsic spin-orbit interaction (SOI), which enables fast electrical manipulation. The SOI of planar quantum dots has mostly been investigated in theoretical studies by examining the SOI already present in the two-dimensional hole gas (2DHG). Here, we study the SOI created by the in-plane confinement by deriving non-perturbative effective Hamiltonians numerically for hole spin qubits. We find that the quantum geometry of the 2DHG naturally emerges, leading to a meaningful non-perturbative definition of pseudospin valid far away from the $Γ$-point. The SOI of the 2DHG and of the in-plane confinement have different forms; therefore, they cannot be turned off simultaneously, ruining the perfect spin-orbit switch functionality of spin qubits. We construct effective Hamiltonians using the symmetry approach for various low-dimensional hole systems: (i) a heavy-hole confined in a SiGe/Ge/SiGe heterostructure, (ii) a light-hole confined in SnGe/Ge, (iii) a gate-defined nanowire in SiGe/Ge/SiGe, and (iv) a hole confined in a Ge/Si core/shell nanowire. The non-perturbative effective Hamiltonians provide results with excellent agreement with the full Hamiltonians.

A Detector Error Model (DEM) is a structured representation of error mechanisms in quantum circuits, which has gained popularity in quantum compilation pipelines for its ability to capture fault-tolerance at a circuit level. It lists error mechanisms as instructions targeting detectors and observables, specifying for each physical fault channel the probability that the fault fires, the detectors it triggers, and the observables it flips. In this paper, we develop an equational theory for DEMs, with its associated categorical semantics. We present a sound, terminating, confluent rewriting system for DEM terms, formulating it as a symmetric monoidal theory (a PROP) over the Giry monad. We prove that every DEM term has a unique normal form, which can be computed efficiently in quasilinear time $O(k|E|\log|E|)$, where $|E|$ is the number of instructions and $k$ bounds the size of a target set. This provides a complete set of invariants (via Tanner graphs) for structural DEM equivalence. We provide the first static decision procedure for DEM equivalence, with rigorous correctness guarantees. It is complete (decides full decoder-equivalence exactly) for non-adaptive quantum error correction (QEC) pipelines, and scales to a sound and applicable decision procedure for partially-adaptive circuits (lattice surgery, distributed QEC, ...) without suffering exponential overhead. We discuss its application to the verification and optimisation of quantum compilers.

Low-frequency gravitational waves ($\lessapprox$ 50 Hz) from core-collapse supernovae are becoming more important for current and future gravitational wave studies. This frequency region is dominated by the global morphology of the explosion and the anisotropic emission of neutrinos from the event. This paper serves as a brief review of both theory and detection (prospects) for gravitational waves in the low-frequency region. We discuss the generation of the linear gravitational wave memory sourced from neutrino emission and show results from an example 15 $M_{\odot}$ Solar metallicity progenitor. We also discuss the detection of the linear gravitational wave memory in current detectors, utilizing a combination of a linear predictive filter and matched templating. Finally we will discuss detection prospects in future detectors such as Cosmic Explorer, Einstein Telescope, the Laser Interferometer Space Antenna, and the Lunar Gravitational-wave Antenna.

We investigate whether relativistic massless classical particles may emerge as the classical limit of massless quanta. To address this question independently of any specific dynamics, environment, or pointer basis, we develop an axiomatic and purely kinematical framework for the coarse-graining approach. In this formulation, a candidate classical phase space is taken as the outcome space of a POVM subject only to minimal classicality and covariance under the relevant spacetime symmetry group. Applying this framework to the Poincaré group, we prove a no-go theorem for massless particles: the covariance requirement is incompatible with the operational conditions for classicality. The theorem leaves open field-like limits of massless quanta, for example the emergence of electromagnetic or gravitational fields, while ruling out classical massless particles, such as classical photons or gravitons.

Tensor networks provide a powerful framework for efficiently representing high-dimensional data and many-body quantum states. Endowing tensor networks with a Riemannian manifold structure provides a natural setting for numerical optimization and analysis. A central feature of tensor networks is their gauge freedom, whose characterisation (captured by so-called fundamental theorems) underlies both their intrinsic structure and the design of numerical algorithms. In this work, we study the interaction between the Riemannian manifold structure and the gauge freedom for several families of tensor networks. Using group actions and Riemannian submersions, we establish a Riemannian fundamental theorem for the tensor network families studied.

Advances in experimental techniques have opened new routes for harnessing non-equilibrium dynamics in mesoscopic quantum systems. In this context, we study the impact of on-site interactions on the transport properties of a continuous quantum thermal machine composed of two coupled oscillators connected to two thermal reservoirs. In the weak system-reservoir coupling regime, where a long-standing debate concerns which reduced description should be preferred, we first show that the Redfield master equation (RME) provides an accurate and unifying framework that interpolates between two well-known limits: the \textit{local} and \textit{global} master equations. By relying on the Hierarchy of Pure States (HOPS), a numerically exact stochastic method, we then explore the full parameter space and show that interactions can be leveraged to tune the efficiency of the thermal machine at high temperatures (while leaving it essentially unchanged at low temperatures), induce non-reciprocal transport under asymmetric reservoir couplings, and generate steady-state entanglement within the junction. We derive expressions for system-bath correlators, such as heat and particle currents, consistently across different frameworks. Our work features on-site interactions to enhance the versatility of quantum thermodynamic junctions and clarifies the role of non-Markovianity and non-linearities in quantum transport.

Dual-species atom interferometry in space promises precise tests of the Universality of Free Fall (UFF), with a sensitivity that grows quadratically with the extended interrogation time accessible in weightlessness. These tests demand exquisite control over the expansion energies of both condensed sources as well as over their differential center-of-mass dynamics. We propose a trap-quenched collimation technique featuring in-trap excitations of collective modes compatible with state-of-the-art atom-chip setups. Using NASA's Cold Atom Laboratory aboard the International Space Station, we demonstrate it on a single-species $^{87}$Rb condensate. By controlling the center-of-mass release dynamics we observe free expansion times up to 700 ms and measure a two-dimensional expansion energy of $k_B \cdot 78\pm 9 \;\mathrm{pK}$ in the imaging plane. A detailed model of the magnetically-induced dynamics indicates that this corresponds to a two-dimensional expansion energy of about $k_B \cdot 15^{+12}_{-5}\; \mathrm{pK}$ along two of the condensate's eigenaxes. Finally, we theoretically study this trap-quenched collimation scheme for a $^{41}$K-$^{87}$Rb mixture, predicting a simultaneous collimation that meets the expansion energy requirements for a state-of-the-art UFF test at the $10^{-15}$ accuracy level.

Higher-dimensional theories of gravity predict that gravitational waves (GWs) can propagate into extra spatial dimensions, leading to modified amplitude damping over cosmological distances. Measurements of GW luminosity distances therefore provide a unique probe of the dimensionality of spacetime. In this work, we constrain higher-dimensional GW propagation using the dark-siren method with the Gravitational-Wave Transient Catalog 4.0 (GWTC-4). We adopt a phenomenological parameterization motivated by braneworld scenarios, in which deviations from General Relativity are characterized by the spacetime dimension number $D$ and a crossover scale $R_c$ governing the transition between four- and higher-dimensional gravity. We perform a hierarchical Bayesian analysis combining 141 compact binary coalescences from GWTC-4 with line-of-sight galaxy information from the GLADE+ catalog. For a prior $H_0 \in [65,77]\ {\rm km~s^{-1}Mpc^{-1}}$ and $\log(R_c/{\rm Mpc}) \in [2.7,4.0]$, we obtain $D = 4.38^{+1.91}_{-1.01}$ (68\% credible interval). We also find that the inferred posterior distribution of $R_c$ accumulates near the upper prior boundary, indicating that the crossover scale remains poorly constrained by current observations. We further show that the inferred constraint on $D$ depends sensitively on the assumed prior range of $R_c$, which determines the characteristic distance scale at which deviations from General Relativity become significant. Our results provide the first GWTC-4 dark-siren constraints on higher-dimensional GW propagation and demonstrate that current observations remain consistent with four-dimensional General Relativity.

The Bose-Einstein condensate dark matter model, where dark matter can be thought of as a non-relativistic, Newtonian gravitational condensate, has recently attracted a great deal of interest. In the present study, we explore the possibility that the dark matter could exist in the form of a self-bound quantum droplet formed by ultradilute quantum Bose mixtures under the action of Lee-Huang-Yang corrections at zero temperature. To this end, we derive an extended equation of state by using the nonrelativistic self-consistent Hartree-Fock-Bogoliubov theory and the hydrodynamic approach. The solutions of the obtained equations of state show that the key parameters of the dark matter halos such as the density, mass, and radius are sensitive to the interspecies interaction and to the quantum fluctuation strength. The stability and the dynamical evolution of the droplet Galactic halos are analyzed by considering small perturbations of the quantum hydrodynamical equations. In order to increase the reliability of our predictions we compare them with some observed data for the Galactic rotation curves.

Superconducting quantum processors could significantly profit from enhanced connectivity together with precise control of interactions and gates between qubits. Here we investigate plaquettes of four qubits that are coupled via a central tunable coupling circuit, so that not only gates between qubits connected by an edge of the plaquette can be executed but also between qubits across the diagonal. By numerically and analytically analyzing parametrically driven processes, we explore $\sqrt{\text{iSWAP}}$-gates between any pair of qubits, also across the diagonal, as well as three-qubit interactions and gates. For experimentally available circuit parameters, we for example find $\sqrt{\text{iSWAP}}$-gates with a gate time of 50 ns and 99.9\% fidelity, which is decreased to 99.4\% if two such gates are executed in parallel on disjoint qubit pairs in the plaquette. For three-qubit gates we find fidelities of 95\% fidelity at a gate time of 200 ns.

We investigate the motion of spinning test particles in the spacetime of a global monopole in the framework of the Mathisson-Papapetrou-Dixon equations. By making use of the symmetries of the spacetime, we obtain a general exact solution to the equations of motion. We show that the particle's trajectories, momenta, and spin can be expressed in terms of three specific functions of the polar and azimuthal angles. We also show that the system is completely integrable. We obtain the general non-geodesic trajectory of the particle, and also examine the particular cases of radial and planar motion. We compare the non-geodesic trajectories of a spinning particle with a non-spinning particle.

The characterisation of quantum phases in strongly correlated systems is a crucial milestone for the deployment of quantum sensors. In this work, we present a Physics-Informed Variational Quantum Classifier (VQC) designed to detect the topological phase transition between the Fermi polaron quasiparticle and the molecular bound state. Unlike conventional Machine Learning approaches, our quantum architecture is constructed via the Trotterised time-evolution of an effective Hamiltonian, ensuring that the learnable parameters correspond to interpretable physical quantities. We show that the VQC efficiently discovers the optimal interferometric protocol, specifically the evolution time and effective bath interactions required to maximise the visibility of Ramsey fringes, thereby clearly distinguishing the Bose-Einstein Condensate (BEC) and Bardeen-Cooper-Schrieffer (BCS) regimes. Furthermore, we report the validation of this classifier on the QRed superconducting quantum processor (BSC-CNS). Despite the intrinsic hardware noise and decoherence, the VQC preserves the relative ordering of the topological phases. We demonstrate that the physics-informed architecture achieves a linear gate complexity $\mathcal{O}(N)$, bypassing the exponential memory wall of classical simulation and ensuring scalability to many-body regimes.

Quantum computing poses a real, broad-based, but bounded and substantially mitigable threat to Bitcoin and Ethereum. We separate the two quantum algorithms that public discussion routinely conflates: Shor's algorithm breaks the elliptic-curve signatures (ECDSA over secp256k1, BLS over BLS12-381) that authorize spending, whereas Grover's algorithm does not meaningfully threaten proof-of-work mining, which is protected by a merely quadratic speedup, fault-tolerant per-operation costs, a square-root parallelization wall, and difficulty adjustment. Folding hardware scaling, the falling resource requirement, a fault-tolerance readiness lag, and expert surveys into a single Monte-Carlo forecast yields a wide, bimodal arrival distribution for a cryptographically relevant quantum computer: about a one-in-six chance by 2035, near 30% by 2040, and about 60% by 2050. Exposure is concentrated and mostly migratable: of Bitcoin's roughly six million quantum-exposed coins only about 2.3 million are irreducibly at risk, while 50 to 65% of Ether sits at key-revealed accounts that can adopt post-quantum signatures. A timely migration beats even an optimistic 2035 machine, so the binding constraint is governance, not technology. A survey of the top twenty cryptocurrencies finds none fully post-quantum. Reproducible models accompany every quantitative claim.

We show that impurity many-body observables can display enhanced sensitivity to ultraviolet deformations of generalized-uncertainty-principle and modified-dispersion-relation type at accessible energy scales. Using a deformed polaron-molecule Hamiltonian constructed to preserve the infrared sector, we quantify the impact of such deformations on spectral and Ramsey observables and implement the corresponding dynamics in a controlled quantum computing setting. We identify regimes near the polaron-molecule crossover where small ultraviolet deformations are strongly amplified, leading to experimentally resolvable changes in quasiparticle properties and spectral response. Our results establish a concrete sensitivity-based route to low-energy quantum-gravity phenomenology in a well-defined many-body platform and delimit the validity of the effective description. Furthermore, we report experimental validation on the QRed superconducting quantum processor (BSC-CNS).

Gravitational waves from close binary white dwarfs will form the bulk of measurements obtained by the Laser Interferometer Space Antenna (LISA). Previous studies have highlighted the importance of including the effects of steady-state mass transfer in waveform models as many individually resolvable white dwarf binaries will be interacting during the LISA observation window. However, few studies have considered the effect of novae on gravitational wave observations and parameter estimations. We fill in this gap by analyzing the detectability of novae in these systems and the biases in physical parameters when these bursts are not considered. We model a nova burst as a rapid loss of mass from the accretor that suddenly shifts the gravitational wave frequency. We analytically predict the signal-to-noise ratios for direct detection of the nova and where bias from mismodeling becomes greater than the statistical uncertainty. We also consider comparison of two halves of future LISA data to detect bursts in a model agnostic way. Our work has implications for identifying and classifying individual nova events as well as constraining the galactic nova rate throughout the entire galaxy, a challenging task for optical surveys.

In a world of various competing quantum computing architectures, hardware-agnostic, full-stack platforms are necessary to bring the full power of quantum computing hardware to domain experts via the cloud. QCI Connect and its Software Development Kit provide a reference architecture for a full-stack platform with a modular design and open-source interface definitions, built to facilitate a community-driven application ecosystem. Here, we present its overall design and features, central interfaces, and lessons learned, both for users of the platform and as a reference guide for future developments.

Accurate and efficient decoding is a crucial component for achieving fault-tolerant quantum computing. Realistic circuit-level noise introduces temporal correlations and degeneracy, making optimal (maximum-likelihood) decoding computationally intractable in general. As a result, practical decoders rely on heuristic approximations, and it is generally difficult to quantify how suboptimal they are, as this strongly depends on the code and noise model considered. In this work, we study the accuracy of practical decoding algorithms under circuit-level noise by comparing them against a maximum likelihood decoding benchmark. Our approach propagates the density matrix through the full memory experiment and computes the optimal decoding decision for each syndrome history. We introduce pruning techniques with rigorous bounds, allowing us to access larger numbers of syndrome-extraction rounds. We apply this framework to small instances of the repetition code and a cellular automaton code, and benchmark minimum-weight perfect matching (MWPM), belief propagation with ordered statistics decoding (BP+OSD), Tesseract, and Planar decoders against optimal decoding. While standard decoders remain close to optimal for the repetition code, we find significant deviations for the cellular automaton code, with BP+OSD deteriorating already in experimentally relevant noise regimes. Moreover, the pruning method developed here highlights that, at low physical error rates, only a narrow fraction of syndrome histories contributes significantly to the logical error rate.

Shallow-depth QAOA on sparse, high-diameter MaxCut instances faces a locality bottleneck: at depth \(p\), local observables can depend only on a bounded neighborhood of the circuit interaction graph. We propose a transport-augmented QAOA that keeps the MaxCut cost Hamiltonian unchanged but enriches the mixer with optimized, unweighted shortcut couplings (scheduled \(XX+YY\)) to collapse the effective interaction-graph diameter. Using exact finite-depth support recursions, we relate optimal shortcut placement to bounded-diameter graph augmentation, and show in benchmarks that (unlike ma-QAOA) performance becomes effectively size-invariant once the diameter is reduced. For bipartite families (base diameter 4), reducing the interaction path to \(d=1\) raises the ensemble-averaged approximation ratio from 0.7378 (ma-QAOA) to 0.9767 at \(p=1\) (\(σ=0.0251\), nine system sizes); on random trees (base diameter 10), at \(p=2\) it improves from 0.9226 to 0.9997 (\(σ=0.0001\)).

Quantum projection noise is the fundamental noise source for the population measurement of spin ensembles. While projection-noise-limited measurements have been extensively studied in atomic systems, corresponding experiments on solid-state spin ensembles remain challenging due to dominant classical readout noise. Here, we report direct measurement of the quantum projection noise of mesoscopic ensembles of nitrogen-vacancy (NV) spin defects at room temperature. Our experiment is enabled by a high optically-detected magnetic resonance (ODMR) contrast of over 20% for a single crystallographic orientation of the defect spins, obtained by combining polarization-selective optical excitation with spin-to-charge conversion. We use our protocol to demonstrate projection noise measurements and spin counting from nanoscale NV ensembles of up to 43 spins. We further demonstrate that the protocol allows for significant gains in sensitivity for magnetometry applications without need for cryogenic operation or high bias magnetic fields.

Quantum information science has traditionally relied on nonclassical resources, such as entangled photon pairs and squeezed states, to achieve measurement performance beyond classical limits. Here, we revisit the multi-pass photonic scheme reported in Nature 450, 393 (2007) to clarify the physical origin of the observed superresolution and the associated claim of supersensitivity. Using a rigorous Jones-matrix formalism, we show that the round-trip evolution of the HQMQ linear optics unit is equivalent to the product of two reflections in polarization space, resulting in an effective rotation operator. This equivalence reveals that the accumulated phase arises from coherent polarization-state rotation on the Poincare'e sphere. The resulting phase accumulation is interpreted geometrically as a progressive realignment of the polarization state during successive forward and backward propagations. To validate the theoretical model, a classical-wave implementation is experimentally conducted, analyzed, and compared with the corresponding Jones-matrix solution. Finally, the scaling behavior of the Fisher information is analyzed to examine the origin of the claimed supersensitivity. The results are further compared with a recently developed coherence de Broglie wavelength framework, which achieves identical superresolution through repeated coherent interactions in a cascaded interferometeric architecture.

Hong-Ou-Mandel (HOM) interference has become a central resource for quantum sensing and metrology owing to its sensitivity to temporal delay and photon indistinguishability. However, existing HOM-based sensing schemes generally rely on inserting a sample into one arm of the interferometer, making the measurement vulnerable to optical loss, alignment instability, and bandwidth-dependent distortion of the interference profile. Here, we demonstrate a symmetry-controlled quantum sensing scheme based on continuous transitions between symmetric (bosonic-like) and antisymmetric (fermionic-like) Bell states in two-photon interference. By imprinting a geometric phase onto the classical pump beam and transferring it to polarization-entangled photons generated via spontaneous parametric down-conversion, we coherently tune the exchange symmetry of the entangled state without altering the temporal or spectral indistinguishability of the photons. The HOM response evolves continuously from bunching to antibunching with a sine square phase dependence, producing a coincidence modulation of approximately 10 * 10^4 counts s^-1 counts/s. In contrast to conventional HOM sensing, the phase-modulation linewidth remains fixed at pi/2, independent of photon bandwidth. Using a birefringent crystal placed directly in the pump beam, we measure thermo-dispersive birefringence with a resolution of the order of 10^{-6} over a broad temperature range. Our results establish exchange symmetry as a controllable resource for robust quantum sensing and symmetry-engineered photonic quantum information processing.

A key challenge of gravitational-wave physics is distinguishing the nature of compact objects involved in binary coalescences, particularly whether they are black holes or neutron stars. Neutron stars are distinguished from black holes by a stronger tidal response, with both static and dynamical aspects directly linked to their rich internal physics. Measurements of the tidal response through gravitational observations constrains the neutron-star equation of state and provides insight into the physics of high-density matter. However, defining the tidal response of neutron stars in general relativity is challenging due to coordinate ambiguities and the complexity of connecting the star's response to binary dynamics and the associated gravitational waveforms. In this paper, we show how the dynamical tidal response of a neutron star can be systematically defined within the worldline effective field theory (EFT) framework, connecting the problem to gravitational-wave scattering off an isolated neutron star. These scattering amplitudes are computed both within the EFT, using standard quantum field-theory techniques, and within stellar perturbation theory (the corresponding ultraviolet theory), where the coupled metric and matter perturbation equations are solved in the stellar interior within general relativity and matched to the analytical Mano-Suzuki-Takasugi (MST) solutions in the exterior. We match the scattering amplitude between effective theory and the ultraviolet theory to obtain the dynamical tidal response. We show the result to be consistent with known expectations, such as the static limit and the behaviour near the neutron star's resonant modes, while also recovering the imaginary part of the dominant oscillation mode induced by gravitational-wave dissipation. We conclude with a discussion of potential future improvements within both the EFT and the perturbation theory.

We consider a scenario for dark matter in the Universe, according to which the dark matter sector is comprised by a dark Standard Model sector which interacts only gravitationally with the ordinary Standard Model sector. This dark Standard Model sector is assumed to have the same symmetries as the ordinary Standard Model, with the couplings and the scale of the mirror Standard Model sector being different than the ordinary Standard Model sector. Specifically, the scale of the mirror Standard Model sector will be assumed to be quite higher compared to the ordinary Standard Model. Also the Yukawa couplings among the mirror Higgs and the mirror fermions are assumed to be different from those of the Standard Model and we examine the effects of the different scale and of the different Yukawas on the evolution of the Universe. As we show, a mirror world phase transition occurs at high temperatures of the baryonic Universe, which can be first order or second order, depending on the scale of the Universe and the Yukawa couplings. These are dark phase transitions which occur quite earlier than the real world Standard Model electroweak phase transition. The case of a second order phase transition is quite interesting phenomenologically, since it can potentially have a direct imprint on the spectrum of stochastic gravitational waves for frequencies probed by the future gravitational wave detectors. Also we examine whether this mirror dark matter world can form atoms and as we show in some scenario the high scale mirror dark matter can have both atomic and subatomic particle components. We also give an approximation of the total equation of state of high scale mirror DM and we discuss how high scale mirror DM can reconcile contradicting observations like the Bullet cluster and the Abell 520 cluster.

The Sachs-Wolfe (SW) effect, arising from large-scale structures in the universe, modifies the frequencies of gravitational waves (GWs) sourced by a cosmological background. We show that for backgrounds with $Ω_{\rm GW}\gtrsim 10^{-10}$, this effect imprints anisotropies and spectral distortions that can be detectable with a network of space-based interferometers (such as LISA + Taiji) and, if not taken into account, may bias the estimate of the theoretical model of the GW background. The effect is particularly enhanced in the high-frequency end of the spectrum. The SW-induced anisotropies and spectral distortions present in a GW background sourced at primordial times will correlate with the SW signature present in the CMB. Any detection of a cross-correlation between the GW anisotropies and the CMB at large scales is therefore a smoking gun for confirming the primordial nature of the background.

We develop a quantum-classical hierarchical equations of motion (QC-HEOM) approach for simulating non-Markovian open quantum systems. The method combines the ensemble-averaged classical path reference of the quantum-classical path integral formalism with a hierarchy of auxiliary quantum influence functionals. By incorporating thermal fluctuations through an ensemble average over reference trajectories, the hierarchy is required to represent only the residual quantum memory associated with the imaginary part of the bath response function. Consequently, unlike conventional hierarchical equations of motion, QC-HEOM does not require Matsubara or Padé expansions of the thermal kernel and exhibits only weak temperature dependence of the hierarchy size. Furthermore, because thermal fluctuations are supplied through reference classical trajectories, the framework naturally extends beyond harmonic baths and enables the incorporation of anharmonic and molecular environments through externally generated trajectories. We derive the formalism and demonstrate its exactness for a harmonic bath. Applications to an asymmetric spin-boson model and the seven-site Fenna--Matthews--Olson complex illustrate the accuracy of QC-HEOM. It reproduces benchmark quasi-adiabatic path integral and hierarchical equations of motion results while requiring substantially fewer auxiliary objects, particularly at low temperatures. These results establish QC-HEOM as an efficient framework for treating residual quantum memory in quantum-classical descriptions of open-system dynamics. The separation of thermal fluctuations from residual quantum memory through the use of Wigner trajectories provides an approximate route toward hierarchical treatments of complex anharmonic environments that are inaccessible to conventional HEOM approaches.

Recent advances in cold atom manipulation enable the study of many-body systems where short-range interactions between neighboring atoms coexist with long-range interactions mediated by photons. Such a combination of interactions makes a theoretical approach challenging beyond mean-field methods. In this work, we develop a neural quantum state based approach to study these systems numerically. We introduce a neural-network architecture capable of handling hybrid Hilbert spaces with large local bosonic dimensions in strongly interacting spin-photon systems. We benchmark this approach on a model of a two-dimensional lattice of Rydberg atoms coupled to a photon mode. The superradiant ground states found in the large spin-photon coupling regime allow us to demonstrate the efficiency of the method in the presence of high photon occupation. Furthermore, the ability to capture spin-spin and spin-photon correlations leads us to observe quantitative deviations in the ground state phase boundaries with respect to mean-field theory. The method extends to other systems with a similar hybrid Hilbert space structure, such as spin-phonon systems, and provides a scalable framework for investigating their ground state properties.

Bayesian confirmation faces a classic problem: where do initial priors come from? In cases with abundant data and repeated updating, different priors tend to converge to the same posterior. However, in frontier research this convergence often fails, and confirmation remains sensitive to priors. We examine how physics practice in the case of gravitational wave research deals with such cases and, normatively, when prior sensitivity should be regarded as epistemically problematic. We offer a practice-based account of the prior problem, so far absent from the philosophical literature on Bayesianism. One upshot is a clearer diagnosis of so-called 'analogue confirmation'.

In this article, we present a novel class of degenerate solutions to the massless Dirac equation, corresponding to a wide variety of electromagnetic 4-potentials and fields, including both zero field and circularly polarized electromagnetic waves. An interesting property of these solutions is that the spin of the particles rotates in synchronization with the electric and magnetic fields of the electromagnetic waves. These results could be utilized for the development of optical memories based on materials supporting massless Dirac fermions, such as graphene.

Optical interferometry is a cornerstone technique for precise phase measurements across various fields. In many applications, for example, biological imaging, it often necessitates stringent limits on light intensity to prevent adverse effects on light-sensitive samples, a condition known as dose-limited regimes. Maximizing the precision per dose is therefore crucial. In quantum metrology, quantum correlations enable high precision in phase estimation while adhering to dose constraints. Nevertheless, photon loss, including absorption by a sample, substantially diminishes the benefits of quantum enhancement in interferometry. In this work, we experimentally investigate a dose-efficient approach to quantum phase estimation using sequential strategies in the presence of loss. Performance of sequential strategies with and without control is evaluated through quantum Fisher information (QFI) per dose. Experimental results show that both sequential strategies exceed the classical limit and outperform the parallel strategy using unbalanced N00N states. Notably, the control-enhanced sequential strategy attains superior QFI per dose, approaching the quantum limit. These results highlight the promise of sequential strategy for imaging and sensing in resource-constrained scenarios, marking a significant step toward practical and efficient quantum metrology in lossy environments.

In this paper, we employ MP codes whose defining matrices are $τ$-optimal defining ($τ$-OD) matrices to construct new quantum codes and quantum $(r,δ)$-LRCs. Specifically, we report the following results: We establish a unified $τ$-monomial decomposition theorem for invertible self-adjoint matrices over finite fields of arbitrary characteristic, which generalizes the result in "Quantum codes using the $τ$-OD MP construction" where the characteristic was required to be odd. Based on this theorem, we prove the existence of $τ$-OD matrices over $\mathbb{F}_{q^2}$ for any characteristic and demonstrate that there exist several new infinite families of $τ$-OD matrices over $\mathbb{F}_{q^2}$ of characteristic $2$. As an application of MP codes involving $τ$-OD matrices, we construct several infinite families of quantum codes with flexible parameters. Within this framework, we present $222$ record-breaking quantum codes that surpass the best-known records maintained in Grassl's database. We propose two effective schemes for constructing optimal pure quantum $(r,δ)$-LRCs via MP codes. Accordingly, we construct four new infinite families of optimal pure quantum $(r,δ)$-LRCs with flexible parameters. Notably, we report an interesting phenomenon by exhibiting $30$ optimal pure quantum $(r,δ)$-LRCs derived from our framework; that is, there exist quantum codes that are not only optimal pure quantum $(r,δ)$-LRCs but also, according to Grassl's database, best-known, optimal, or record-breaking quantum codes. To the best of our knowledge, the new discovery that quantum codes are simultaneously optimal pure quantum $(r,δ)$-LRCs and record-breaking quantum codes has not been previously reported in the literature.

We explore soliton interactions in a homogeneous spinor $F=1$ Bose Einstein Condensate (BEC) in the presence of a magnetic field, focusing on dark bright dark and bright dark bright configurations. We investigate how these interactions depend on the phase differences among bright solitons and their influence during the dynamics. Our findings align with prior non spinor results, i.e., repulsion among in phase bright solitons and attraction among out of phase pairs in self repulsive atomic BECs. The potential bright soliton attraction, added to the short range repulsion of dark dark soliton interactions, can lead to bound states. However, we find that these bound states break in the presence of spinor interactions due to the particle exchange dynamics between the hyperfine states of the components. Additonally, we develop an effective classical model to describe the soliton dynamics, using a Lagrangian approach. The accuracy of the model is tested by comparing it against numerical simulations. Our results suggest that the proposed model captures the essential features of soliton behavior in the presence of spin interactions, and provides congruent soliton trajectories and interspecies particle exchange dynamics in most of the cases.

Quantum walks provide a versatile framework for quantum algorithms across a wide range of applications. We develop efficient classical simulation methods for Szegedy quantum walks that avoid explicit construction of the full unitary evolution operator. Unlike previous approaches restricted to a particular walk formulation, our framework is built from fundamental update and reflection operators, enabling the simulation of a broader class of Szegedy walk formulations. We further extend these methods to phase-estimation-based algorithms coupled to the walk, including implementations suitable for large sparse graphs. The resulting methods achieve optimal $O(N^2)$ complexity for dense graphs with $N$ nodes. For sparse graphs, the computational cost scales linearly with the number of edges, which is $O(N)$ in many cases. We implement the framework in the Python package SQWLib and illustrate its capabilities through simulations of representative algorithms, including quantum simulated annealing and quantum search on graphs. These results provide a practical tool for studying Szegedy-walk-based algorithms numerically beyond purely analytical treatments.

Fourier analysis has become a crucial tool for understanding the expressivity of Variational Quantum Circuit (VQC) models, as well as an important indicator of barren plateaus (BP). While existing literature has only studied angle-embedded VQCs in a noiseless environment, here we develop the Fourier analysis of VQCs with non-linear data embedding, with particular focus on amplitude embedding, which provides a naturally compact encoding scheme. We first investigate a subtle difference in the domain of input features within amplitude embedding that leads to a distinct expressivity of the zero-frequency Fourier coefficient. By assuming that the ensemble of unitaries generated from the parameter space forms at least a 2-design with respect to the unitary group, we derive, via Weingarten calculus, that the mean of the Fourier coefficients is concentrated at zero, and the variance scales at an exponentially decaying order with respect to the multi-dimensional frequency magnitude. When a noise channel with unitary Kraus operators and probabilities $\{p_k\}$ is taken into account, the variance is further suppressed by a factor $\left(\sum_k p_k^2\right)^{Q}<1$, where $Q$ is the number of channel instances applied. Furthermore, we demonstrate and validate the analytical results through simulations, both noiseless and noisy, including a case where target functions are decomposed into non-integer frequencies, highlighting the practical utility of the approach. Our results establish a rigorous Fourier framework for amplitude-encoded VQCs, offering both theoretical guarantees on expressivity, hence trainability scaling in the frequency domain, as well as practical simulations for deployment on noisy quantum devices.

We study the simultaneous estimation of partial-transpose moments $p_j(ρ_{AB})=\mathrm{Tr}[(ρ_{AB}^{T_B})^j]$, $j=2,\ldots,K$, of an unknown bipartite $n$-qubit state from independent copies under an explicit active-memory constraint. We give a sequential qubit-reuse realization of the partial-transpose permutation that uses at most $2n+1$ active qubits, independent of $K$, and estimates all moments $p_2,\ldots,p_K$ to uniform additive error $ε$ with total copy complexity $O(K\log K/ε^2)$. We also prove two converse bounds. First, any uniformly accurate simultaneous estimator requires $Ω(K/ε^2)$ copies in the worst case. Second, the same scaling holds on an explicit isospectral two-qubit negative-partial-transpose (NPT) family whose ordinary moments are constant while the partial-transpose moments vary. These results characterize the copy complexity of the partial-transpose moment hierarchy up to a logarithmic factor and extend simultaneous nonlinear-functional estimation from ordinary state powers to partial-transpose spectral data under active quantum memory independent of the target moment order.

Next-generation gravitational-wave observatories will observe binary neutron-star mergers with much higher signal-to-noise ratios, over much longer durations and across broader frequency bands than current detectors. These long-duration signals present a major computational challenge for Bayesian parameter estimation. Reduced-order quadrature is a promising approach for accelerating inference, but in this regime its standard construction encounters severe memory and accuracy limitations. We present a practical reduced-order quadrature construction for long binary neutron-star signals with time-dependent detector response and full effects of the observatories' free-spectral range. Our approach combines improved adaptive frequency sampling, disk-backed streaming, and subbanded reduced-order quadrature construction, enabling efficient and accurate reduced-order models for signals that were previously intractable. We demonstrate for the first time reduced-order Bayesian inference on an approximately 2 h binary neutron-star signal extending down to 5 Hz and with signal-to-noise ratio 2090. We show the resulting reduced-order quadrature remains sufficiently accurate for practical inference. The full analysis is carried out in about 48 h using 128 CPU cores. We also find that, when time-dependent detector-response effects are included, a single Cosmic Explorer detector can localize such a signal to a 90% credible sky area of approximately $41~\mathrm{deg}^2$, with important implications for multimessenger astronomy and cosmology. These results demonstrate that reduced-order methods can make next-generation binary neutron-star inference computationally feasible.

We present the propagation of massive Majorana neutrinos crossing an exact gravitational sandwich wave background. By employing a Takagi factorization to decouple the matrix field equations, We show that the localized pulse breaks the kinematic alignment of the mass eigenstates as they propagate through the wave zone, leaving a permanent residual phase shift across the transverse profile of the wavefront. The wave asymmetry couples directly to the Majorana spin tensor, forcing a transition from left-handed to right-handed configurations. Behind the wave tail, the residual drift parameters act to polarize the state's chirality, introducing a strong anisotropy across the transverse momentum plane. These results show that a passing gravitational wave leaves a modification on both the flavor and helicity of a neutrino beam, as a sign of the gravitational memory effect.

We investigate the decoherence of a quantum spatial superposition of a static particle in Schwarzschild and Reissner-Nordström black holes. By treating the particle as a localized classical source coupled to a quantum scalar field, we reformulate the decoherence process in the Danielson-Satishchandran-Wald (DSW) gedankenexperiment through coherent state generation and derive the local expression for the decoherence functional in terms of the Wightman function. In the long-time limit, the decoherence rate is shown to be characterized by the low-frequency behavior of the Wightman function. We then employ the asymptotic matching method to calculate the analytical expressions of the Wightman functions in the Boulware, Unruh, and Hartle-Hawking vacua. We show that the decoherence behavior depends on the quantum state of the environmental field. While the Boulware vacuum gives vanishing decoherence for a static superposition, the thermal effects associated with Hawking radiation in the Unruh and Hartle-Hawking vacua can induce nonvanishing decoherence.

We study a system qubit repeatedly interacting with the same environmental qubit, with a reservoir acting on the environment between collisions via a completely positive, trace-preserving map. We show that complete suppression of system--environment correlations uniquely requires a full environmental reset, recovering a semi group dynamics with a time-independent Gorini--Kossakowski--Sudarshan--Lindblad generator, whereas a partial reset yields a continuous transition between Markovian and non-Markovian regimes governed by a single dimensionless relaxation parameter. For a resonant excitation-exchange interaction, we obtain exact closed-form expressions for the Bloch-vector dynamics for both a generalized depolarizing channel and a generalized amplitude-damping channel acting as the reservoir-induced map. Using the Breuer--Laine--Piilo measure and a Choi-matrix CP-divisibility witness, we identify three distinct dynamical regimes across the parameter space: CP-divisible Markovian dynamics, CP-indivisible but P-divisible dynamics, and non-P-divisible non-Markovian dynamics. The boundaries between these regimes, and the structural differences between uniform and anisotropic environmental relaxation, are characterized numerically.

We study quantum gravitational effects inside hyperbolic black holes with both outer and inner horizons by solving the Wheeler--DeWitt (WDW) equation in the minisuperspace approximation. The WDW equation contains a tachyonic region where the effective potential becomes negative. We develop a numerical method that consistently evolves the wave function across this region and obtain stable solutions throughout the entire minisuperspace. For small values of the parameter $κ$, which controls the strength of quantum gravitational effects, the wave packet propagates along the classical trajectory with only mild quantum spreading. As $κ$ increases, enhanced quantum effects lead to significant spreading of the wave packet during its propagation. Nevertheless, when the initial state is localized near the outer horizon, the wave packet becomes localized again in the vicinity of the inner horizon. We refer to this recovery of localization as a refocusing phenomenon. This result suggests that, if the geometry is classical near the outer horizon, it becomes classical again near the inner horizon. Within the minisuperspace approximation, inner-horizon formation is not obstructed by quantum gravitational effects.

We investigate the hydrodynamic properties of a strongly coupled non-Abelian plasma dual to a four-dimensional AdS black brane with a nonminimal coupling of the form $q_2 F^{(a)αβ}F^{(a)γλ}R_{αγ}R_{βλ}$ in the bulk action. This higher-derivative term introduces a direct interaction between the Yang-Mills field strength and the Ricci tensor, leading to corrections beyond the minimal Einstein-Yang-Mills theory. Using a perturbative expansion in the small coupling $q_2$, we construct the black brane solution up to first order and employ the fluid-gravity correspondence to compute two key transport coefficients: the DC color conductivity $σ$ and the shear viscosity to entropy density ratio $η/s$. In holography, $η/s$ is inversely proportional to the square of the coupling constant of the boundary theory, while $σ$ in Einstein-Maxwell theory satisfies a universal lower bound $σ\ge 1$ (in units where $\hbar=1$), saturated for uncharged black holes and characterizing a perfect quantum critical fluid. Our results reveal that the nonminimal coupling significantly alters these transport quantities. For the DC conductivity, we find $σ= 1 - q_2 \bigl(9κQ^2/(L^2 r_h^4) + 7κ^2 Q^4/(4 r_h^8)\bigr)$, indicating a violation of the conductivity bound for $q_2>0$ while the bound is preserved for $q_2<0$. For the shear viscosity, we obtain $η/s = (1/(4π))\bigl(1 + q_2\, 7κ^2 Q^4/(2 r_h^8)\bigr)$, showing that the KSS bound is modified by a term linear in $q_2$. The sign of $q_2$ determines whether the ratio increases above or decreases below the universal value $1/(4π)$. These findings highlight the sensitivity of holographic transport to curvature-coupled gauge interactions and provide a controlled example of how higher-derivative corrections influence the hydrodynamic regime of strongly coupled plasmas.

Near-term variational classifiers incur substantial error and latency from two-qubit gates, yet practitioners often assume that additional entangling depth is the default route to higher accuracy. This work studies Optimal Quantum Measurement Decoding (OQMD): optimizing how quantum outcomes are mapped to classical labels by training a readout layer before measurement, jointly with the variational circuit, without adding CNOTs. Experiments use trainable triple single-qubit rotations as one concrete, hardware-native realization of OQMD; other single-qubit parametrizations fit the same classical outer loop. On the Iris benchmark with a 30-point stratified test split, the best observed 0-CNOT configuration with OQMD reaches 83.33\% accuracy, with a 96\% at 9 CNOTs, exceeding the best 18-CNOT controls (56.67\%) and the best 18-CNOT configuration with OQMD (66.67\%) under a common protocol. A six-point CNOT-depth series from 0 to 18 (fixed optimizer, iteration budget, random-seed count, and ZXZ readout) shows that the highest raw scores need not occur at the largest template, so aggregate complexity is not summarized by CNOT count alone. Because run-level accuracies are discrete and non-Gaussian, we emphasize best-observed scores and, where a global comparison of pooled runs is required, Mann--Whitney $U$ tests rather than parametric tests on means. Across architectures, OQMD shows statistically consistent but magnitude-dependent gains: large peak lifts on minimal circuits coexist with a small pooled mean shift on complex 18-CNOT runs ($p\approx 0.03$) that is not ``universal'' in the sense of uniformly large practical effects.%

Precise timing synchronization is essential for distributed quantum networking, enabling entanglement distribution, quantum teleportation, and entanglement swapping across remote nodes. Existing synchronization architectures rely on dedicated timing-distribution infrastructure, most notably White Rabbit networks, which constrain topology, scalability, and deployment in free-space and satellite environments. Here we demonstrate link-free synchronization of quantum network nodes using independently operating miniature rubidium atomic clocks and computational post-processing. We validate the approach on a deployed metropolitan-scale telecom fiber network spanning three geographically separated nodes. Following drift correction, atomic-clock-based synchronization achieves timing performance approaching that of a White Rabbit benchmark and remains stable over continuous 8-hour operation. As a stringent test of quantum-network functionality, we observe Hong-Ou-Mandel interference across spatially separated nodes with visibility exceeding 70%, statistically equivalent to that obtained using dedicated White Rabbit timing links. To the best of our knowledge, this represents the first observation of quantum interference across a deployed metropolitan-scale telecom fiber network synchronized entirely without dedicated timing-transfer infrastructure. These results establish atomic-clock-based synchronization as a scalable, topology-independent alternative to conventional timing-distribution architectures and a practical pathway toward terrestrial, airborne, and space-based quantum networks where dedicated timing links are unavailable.

In this paper, we show that under a suitable set of symmetry arguments and asymptotic conditions, the uniqueness of the Kerr solution can be proven. None of the conditions considered herein explicitly assume the validity of the Einstein equations. Hence, we are able to construct a theory agnostic uniqueness theorem for the Kerr spacetime. This result has implications for theories of quantum gravity or modified theories of gravity which wish to excise singular behaviour from their corresponding spacetimes. Furthermore, this work is complimentary to Penrose's singularity theorem. While the spacetime considered herein is not as general as that assumed by Penrose, we show that singularities can be unavoidable even when the validity of the Einstein equations is not presupposed.

Estimating the sensitivity of gravitational-wave searches is important for a wide variety of scientific applications spanning astrophysics and fundamental physics. In this work, we develop a fast semianalytic approximation for estimating matched-filter search sensitivity to physical effects not explicitly modeled in the template bank. This approximation utilizes fitting factors, i.e., the maximum overlap of a candidate signal over the search bank. As illustrations, we compare our estimates to the actual performance of searches against spinning binary neutron stars, and evaluate search sensitivity to compact binaries possessing orbital eccentricity or deviations from general relativity. Our work thus paves the way for fast sensitivity estimates for a variety of applications, including unmodeled effects in template banks such as deviations from general relativity, environmental effects, gravitational lensing, and waveform/calibration systematics.

In this work, it is shown that many quantum error-manipulating techniques, such as distillation, error mitigation, and dynamical decoupling, are special cases of the most general framework for quantum error correction. This unifying perspective is achieved by extending quantum error correction to include state-adaptive and channel-adaptive settings, as well as multi-stage coding scenarios. Based on this insight, a model of self-correcting quantum memory is also proposed. This work clarifies the relationship among these techniques and illustrates, through explicit constructions, how the unified perspective can guide the design of reliable quantum information systems.

We study black hole thermodynamics using a two-dimensional effective theory obtained by dimensional reduction of four-dimensional Einstein--Maxwell theory. For spherically symmetric charged black holes, the resulting dilaton gravity has a nonlinear potential that reproduces the semiclassical phase structure of four-dimensional AdS black holes, including the Hawking--Page transition and the small/large Reissner--Nordström--AdS black hole transition. This shows that the two-dimensional theory before taking the near-horizon and near-extremal limits captures non-extremal thermodynamics beyond the Jackiw--Teitelboim gravity regime. We also include electromagnetic Kaluza--Klein modes on the internal sphere and integrate them out to derive the one-loop effective dilaton gravity. At leading order in the derivative expansion, these corrections appear as constant shifts in the black hole entropy and in the effective charge parameter of the dilaton potential. Therefore, the semiclassical phase structure is not qualitatively modified within this leading local approximation.

We investigate the viability of scalar-tensor (quintessence) models of dark energy with a quartic polynomial nonminimal coupling to gravity and a linear scalar potential. The polynomial nonminimal coupling is used to ensure that the field is well stabilized in the early universe. We perform a systematic exploration of the parameter space spanned by the quadratic and quartic nonminimal couplings, as well as the slope of the potential. We confront the predictions of the model with the latest Dark Energy Spectroscopic Instrument (DESI) constraints on the dark energy equation of state and also with complementary bounds from local tests of gravity, including solar system constraints and limits on the time variation of the effective Newton's constant. We identify bands in parameter space where all these constraints are satisfied, finding such bands to be very narrow.

This work presents an experimental setup for implementing two-qubit operations on neutral atoms ($^{87}$Rb) with the possibility of using two different Rydberg excitation schemes. One of them uses 5P$_{1/2}$ as the intermediate level and applies the second-stage beam locally to the addressed atoms. The second scheme uses the 6P$_{3/2}$ level; in this scheme, the particles to be entangled are moved to a separate zone through which both Rydberg beams pass. The advantages and limitations of both schemes are analyzed. Based on numerical modeling performed with a Julia package developed by the authors, it is demonstrated that the spatial configuration has a greater effect on quantum-operation fidelity than the choice of intermediate level. An experimental implementation of the scheme using the 6P$_{3/2}$ level is demonstrated, making it possible to achieve a two-qubit operation fidelity of 94%.

We present a custom-designed ultrahigh vacuum (UHV) cluster tool developed for studying shallow nitrogen-vacancy (NV) centers in diamond, enabling in situ diamond surface preparation, characterization, and single NV center dynamics measurements within a single connected platform. The system combines a surface science chamber for controlled surface modification and analysis with a cryogenic confocal microscope chamber dedicated to NV spin and optical measurements. This integrated approach enables a direct correlation between diamond surface chemistry and the resulting NV spin and charge properties. The instrument provides a versatile platform for systematic studies of surface-induced decoherence mechanisms and charge dynamics for shallow NV centers, and establishes a pathway toward reproducible surface engineering for quantum sensing applications.

Electron transfer through chiral structures can exhibit spin asymmetry, known as the chiral-induced spin selectivity effect, whose microscopic origin remains an open question. While path-interference within the chiral moiety has been proposed as a key mechanism, its experimental validation requires precise and versatile tunability of system parameters. Here we implement a programmable quantum simulation of spin-dependent electron transfer in a donor--chiral-bridge--acceptor model using a trapped ion. The bridge is encoded in internal states of the ion with tunable nearest- and next-nearest-neighbor couplings, while donor and acceptor states are coupled via a spectator bosonic motional mode. We observe spin-dependent interference within the bridge, and further reveal spin-dependence in donor-to-acceptor transfer dynamics, controlled by amplitude and phase of the coupling parameter. Our results identify interference among spin-dependent pathways as a microscopic origin of spin-dependent transfer, and open a route toward quantum simulations of complex chiral lattices with multi-level and bosonic degrees of freedom.

Compact graphs are graphs for which the fractional automorphism polytope has no genuinely fractional vertices. This paper proposes a quantum analogue of this idea by evaluating the fundamental magic unitary of the quantum automorphism group on states, which we show to produce a closed convex set of doubly stochastic matrices sitting between the classical automorphism polytope and the full fractional automorphism polytope. Our main result is that the natural quantum analogue of compactness is classical, that is, a quantum compact graph is classically compact. We also relate this set to the quantum orbital algebra and obtain a hierarchy of classical and quantum compactness pseudo notions. The framework recovers familiar consequences of compactness through commutants and suggests quantum analogues of generous transitivity and distance-transitivity. We also isolate examples and open problems indicating where quantum symmetries may strictly refine the classical compactness theory.

We investigate several astrophysical motivated properties of a four-dimensional black hole solution in the framework of $f(R)$ gravity. The model is characterized by a single additional parameter, $α$, which encodes nontrivial deviations from General Relativity. Using this black hole spacetime as the background geometry, we analyze: (i) the quasinormal modes of massless scalar perturbations (employing several complementary methods), and (ii) the greybody factors associated to the propagation of massless test scalar fields. Regarding quasinormal modes, we study the response of black holes to massless scalar perturbations using three independent approaches: the well-established sixth-order WKB semi-analytic method, analytic expressions, and the analytical expression in the eikonal limit. We examine the behavior of both the real and imaginary parts of the quasinormal frequencies as functions of the parameter $α$, the overtone number $n$, and the multipole number $\ell$. In addition to that, we compute the greybody factors within the WKB approximation for various combinations of the relevant parameters. In particular, we focus on the propagation of test massless scalar fields and investigate how the greybody factors depend on the multipole number $\ell$, and the free parameter $α$. The impact of the aforementioned parameters on the absorption cross-section and the reflection and transmission coefficients is studied in detail.

The ringdown phase of a binary black hole merger encodes fundamental information about the remnant through its quasinormal mode (QNM) spectrum. Extracting multiple modes from gravitational-wave data is essential for black-hole spectroscopy but remains challenging due to the short duration of the signal, limited signal-to-noise ratio (SNR), and interference between modes. In this work, we investigate the applicability of Empirical Mode Decomposition (EMD) and Variational Mode Decomposition (VMD) to the analysis of post-merger gravitational-wave signals. Using Monte Carlo simulations of noisy ringdown signals composed of pairs of QNMs, we assessed the ability of these methods to separate the two modes and estimate their frequencies. The instantaneous frequency is obtained via the Hilbert transform (HT) and our new proposed method, named instantaneous frequency and amplitude determination (IFAD). We analyze performance over a wide range of SNR values relevant to current and future gravitational-wave detectors. Our results show that both EMD and VMD can resolve multiple modes, and VMD provides significantly more accurate frequency estimates. We also introduce a modified instantaneous frequency estimator that improves accuracy over the Hilbert transform. The study quantifies the conditions under which two-mode resolution is feasible and highlights the limitations imposed by closely spaced mode frequencies and short signal duration. These results are relevant for current observations and for future high-SNR detections expected from space-based and next-generation ground detectors.

We show that the equal-prior Helstrom trace-distance geometry of qubit states is recovered from Connes spectral distance in a finite scalar-qubit-scalar model. The two scalar reference sectors couple isotropically to the qubit block through identity Dirac links, so that the full Bloch ball, including mixed states, inherits its standard chordal trace-distance geometry from the finite spectral metric. The scalar-sector distances serve a distinct calibration role: they determine the individual link lengths, satisfy a Pythagorean consistency relation, and reconstruct the middle-sector scale.

We revisit the transition from inspiral to plunge for extreme mass-ratio inspirals on quasi-circular, inclined orbits in Kerr spacetime from the perspective of catastrophe theory. Our goal is to uncover the mathematical structures underlying the universality of the transition dynamics, which remains governed by the same Painlevé I differential equation as for equatorial inspirals despite the additional complexity. We first analyze the solution of the Painlevé I equation selected by the physical boundary conditions of slowly evolving quasi-circular inspiral at early times. We argue that these conditions uniquely select the tritronquée solution of Painlevé I. We then compare existing high-accuracy analytic approximations of the tritronquée solution with direct numerical integrations of the Painlevé I equation, finding comparable accuracy and improved stability under differentiation and integration for the analytic solution. In the second part of this work, we show that the equilibrium structure of the Kerr radial effective potential admits a natural interpretation in terms of catastrophe theory. Equatorial orbits are associated with the fold catastrophe, while inclined orbits are described by the cusp catastrophe. In both cases, the transition to plunge corresponds to slow evolution across fold lines of the catastrophe manifold, providing a geometric explanation for the universal appearance of the Painlevé I equation in the transition dynamics.

We investigate a representative set of static solitonic solutions of the Einstein-Proca theory in the Newtonian regime, where the field frequency approaches the particle mass, $ω\to μ$, and compare them with the corresponding solutions of the spin-1 Schrödinger-Poisson system, which provides the effective description in this limit. While this correspondence is relatively straightforward in the Einstein-Klein-Gordon case, the vector nature of the Proca field, combined with the enhanced $U(3)$ symmetry of the nonrelativistic spin-1 regime, gives rise to several nontrivial features that require careful analysis. We establish a mapping between the two descriptions by identifying $\ell=0$ electric Proca stars with radially polarized (hedgehog) configurations and $\ell=1$ electric Proca stars with linearly polarized configurations. We further clarify some aspects of the ground state and resolve several apparent discrepancies between relativistic and Newtonian solutions, particularly concerning their morphology and stability properties. An important conclusion of this work is that the nonrelativistic regime supports a richer spectrum of stable equilibrium configurations than the relativistic theory, including stable excited states.

We argue that the gravitational path integral for rotating black holes is periodic in the angular velocity, implying the quantization of angular momentum in arbitrary dimensions for either asymptotically flat or AdS boundary conditions. In AdS$_3$, this periodicity is a consequence of the boundary mapping class group. In higher dimensions, the periodicity arises from an infinite family of saddles labeled by integer shifts of the angular velocity unrelated to the boundary mapping class group; summing over these saddles enforces quantization independently of any large boundary diffeomorphism. We construct these saddles explicitly for Kerr-Newman black holes in both asymptotically flat space and AdS$_4$, and observe that even the path integral for the 4D Schwarzschild black hole, typically the simplest case, receives contributions from an infinite set of rotating saddles.

We develop an ambitwistor-space formulation of conserved spinning correlators in four-dimensional CFTs. We show that conformal symmetry and conservation are trivialised by the ambitwistor Penrose transform, and construct the two- and three-point ambitwistor correlators explicitly. We further propose a simple formula for projecting onto parity-even and parity-odd sectors, and verify it in several non-trivial examples. The resulting correlators furnish a natural basis compatible with a boundary double copy for arbitrary spin. At three points, we show that they localise on degree-three curves in a two-twistor representation. We also explain how the formalism incorporates boundary propagators for unitary operators of integer conformal dimension, and extend it to conformally coupled scalars.

Construction of physically realistic theories admitting regular black hole solutions remains an important open problem in gravitational physics. While theories with electromagnetic fields minimally coupled to gravity have been extensively studied over the past two decades, theories with nonminimal couplings remain comparatively unexplored. We investigate theories containing the interaction terms $R F_{ab}F^{ab}$, $R_{ab} F^a_{\ \, c} \, F^{bc}$ and $R_{abcd} F^{ab} F^{cd}$, which generically arise in low-energy effective Lagrangians. We prove that magnetically charged regular black holes are excluded, except possibly for finely tuned choices of coupling constants, and argue that a similar conclusion applies to electrically charged regular black holes. We further show that similar conclusions hold for Lagrangian terms of the form $f(R,F_{ab}{\star F}^{ab})$.

Analog quantum simulation provides a highly controlled platform to study diverse quantum many-body phenomena. However, current methods for state initialisation are limited to thermal ensembles or uncorrelated product states. Here we present a hybrid approach that complements analog preparation with a digital quantum-gate protocol. This approach enables the engineering of target states with specific, long-range spin-correlations from the same initial resource state. By applying collisional gates to adiabatically prepared and filtered four-fermion singlet chains, we program diverse spin-correlation patterns, including that of a Heisenberg chain. We measure the spin correlations using a sequence of quantum gates followed by singlet-pair measurements. Our method paves the way to the targeted preparation of strongly correlated states of matter.

Establishing realistic expectations for the dark matter (DM) distribution near a supermassive black hole (BH) is essential for assessing environmental imprints on gravitational wave (GW) signals. Using the Galactic Center as an observationally constrained case study, we investigate the evolution of DM density spikes under gravitational perturbations from the nuclear stellar and black hole populations surrounding the central BH. We find that scattering of DM particles by the nuclear star cluster depletes the DM distribution at radii $r \sim 10^{-1}~\mathrm{pc}$, far outside the region where the relevant GW signals are produced. At smaller radii, at $r \sim 10^{-3}~\mathrm{pc}$, the closest known stellar perturbers, S2 and S38, induce only negligible changes to the density profile. At still smaller radii, where no stellar perturbers are currently known, we assess the cumulative impact of past EMRIs by modelling successive mergers with stellar-mass BHs of mass $\sim10~\mathrm{M}_\odot$, in numbers consistent with the expected rate over $10~\mathrm{Gyr}$. We find that these events do not erase the central overdensity, reducing the density only to approximately $82 \%$ of its initial value at $r \lesssim 10^{-5}~\mathrm{pc}$. Our results indicate that, at least at the Galactic Center, DM overdensities around the central BH are expected to remain largely intact under stellar perturbations and plausible stellar-BH merger histories.

The performance of superconducting integrated circuits for quantum applications is fundamentally limited by material-related losses. Tantalum, as an emerging material for next-generation quantum circuits, has attracted considerable attention in recent years after demonstrating breakthrough performance in both superconducting microwave resonators and qubits. Concurrently, a growing body of work is devoted to the operation of tantalum-based circuits and related fabrication techniques. This interest is further stimulated by tantalum thin films polymorphism resulting in a variety of its crystalline structure, superconducting properties, coherence, etc. Furthermore, tantalum circuits exhibit distinctive features in cryogenic experiments, which have not been observed in aluminum- or niobium-based ones. In this review, we summarize the recent research of tantalum thin films growth and phase selection mechanisms on various substrates, key aspects of fabrication and performance of superconducting circuit, including a material first-principles theoretical study. In conclusion, we address a number of open issues, including the role of \b{eta}-phase impurities, the effect of hydrofluoric acid solutions on chain characteristics, and the anomalous behavior of α-tantalum chains at cryogenic temperatures.

The Laser Interferometer Space Antenna (LISA) will detect gravitational waves (GWs) from dozens of extreme- and intermediate-mass-ratio inspirals (EMRIs/IMRIs). These sources will stay in-band for months to years, offering extraordinary scientific potential. However, their fully phase-coherent analysis standard in current pipelines imposes stringent waveform accuracy requirements; failing to model the signal over such long durations can result in significant systematic biases. To address this, we formulate a robust semi-coherent Bayesian inference framework that segments the data into independent blocks, analyzes each block coherently, and recombines the results incoherently. By restricting phase-tracking to much shorter block durations, this approach prevents significant accumulation of phase errors. We implement this methodology in \textsc{SPLIT} (Semi-coherent Posteriors for Long-Inspiral Templates), a GPU-accelerated Python package. Applying \textsc{SPLIT} to an environment-rich injection, we demonstrate that while a fully-coherent vacuum-GR analysis incurs a maximum 1D systematic bias of $\approx 4.8σ$ from the truth, the shorter integration window of our semi-coherent approach restricts such biases to $\lesssim 0.5σ$. Overall, despite a fractional loss of optimal signal-to-noise ratio, the substantial improvement in parameter accuracy offered by the semi-coherent approach presents a highly advantageous trade-off for LISA and other future GW detectors.

We develop a Mars-centered post-Newtonian framework for relating barycentric coordinate time, Mars-centered coordinate time, a conventional Mars surface time scale, and the proper times of landed and orbiting clocks. The construction follows the International Astronomical Union BCRS/TCB formalism, introduces Areocentric Coordinate Time (TCA), and writes each clock transformation as a secular rate plus zero-mean periodic terms. Terms are retained when their fractional-frequency amplitude exceeds 5e-18 or their one-way accumulated timing amplitude exceeds 0.1 ps. The numerical realization uses the GMM-3 Mars gravity field through degree and order 120, point-mass tides from the Sun, Phobos, and Deimos with origin and dipole terms removed, and bounds on omitted local c^-4 and external-perturber terms. Representative low-Mars-orbit, areostationary, Phobos-/Deimos-distance, and highly elliptical relay regimes are evaluated. Relative to the adopted Mars surface scale, a 300 km near-polar clock is slower by 4.56 microseconds per day, while areostationary and Deimos-distance clocks are faster by 9.13 and 9.52 microseconds per day. The leading Mars-J2 timing line is about 87 ps at 300 km altitude and remains several ps near areostationary radius for inclined or librating areosynchronous cases; perihelion-scaled solar tides become retained sub-ps terms in high relay orbits. The result is a reference-system and model-retention framework, not a final operational Mars Time Ephemeris. A realized sub-ps system still requires a selected planetary ephemeris, Mars orientation and seasonal-gravity model, spacecraft orbit determination, calibrated link delays, and covariance analysis. Time-variable low-degree gravity from seasonal CO2 exchange is a leading surface-realization term and must be modeled, monitored, or empirically bounded before sub-ps Mars surface-scale claims are made.

Galaxy rotation curves provide a powerful probe of mass distributions in spiral galaxies. We present a general relativistic analysis of NGC 7331 using its observed rotation curve and WISE W1 (3.4 $μ$m) photometry to constrain the stellar mass profile. Adopting a static, spherically symmetric spacetime with anisotropic matter (vanishing radial pressure), we fit a modified exponential azimuthal velocity law to kinematic data, reconstructing metric functions and deriving enclosed mass, energy density, and tangential pressure profiles. The stellar mass underpredicts the total gravitating mass at intermediate-to-large radii, indicating dominant dark matter. The physical viability of the resulting relativistic model is examined through energy conditions, causality constraints, and the stability of circular orbits. A comparison with the standard Navarro--Frenk--White dark matter profile has been done. Overall, this study demonstrates that combining rotation-curve data with photometric stellar mass estimates within a general relativistic framework provides a consistent and physically viable description of the mass distribution in spiral galaxies such as NGC~7331.

We study memory effects in quantum cosmology by extending the semiclassical Wheeler-DeWitt framework beyond its usual local form. The main idea is to introduce a causal memory kernel at sub leading order, rather than imposing fractional derivatives directly by hand. In this setting, fractional time evolution appears as an effective description of the underlying nonlocal dynamics. We apply the framework to cosmological perturbations in de Sitter space and find a correction to the primordial power spectrum with a characteristic $k^{3/4}$ scaling. This contribution mainly affects high $l$ CMB temperature anisotropies, in contrast with standard semiclassical quantum gravitational corrections, which are strongest at large angular scales. We also discuss how the same memory-dependent dynamics may affect primordial non-Gaussianity, producing scale dependent corrections to the bispectrum and possible deviations from the usual squeezed limit consistency relation. Since the memory coefficient controls short scale power, it may also influence structure formation and could require some tuning in order to give phenomenologically acceptable astrophysical environments. Finally, we suggest that a cyclic extension of the Hawking-Hartle no-boundary proposal may provide a setting in which the effective memory strength can evolve across successive cosmological histories. In this way, the framework gives a concrete realization of fractional quantum cosmology based on memory effects and also points to possible observational signatures of nonlocal quantum gravitational dynamics.

New flavour non-abelian gauge groups, which may arise as part of a fundamental theory of flavour, can lead to distinctive flavour-transfer processes. When restricted to the lepton sector, such processes partially mimic the standard charged current interactions at low energy. We explicitly study such constructions with various flavour structures, and investigate systematically all relevant accelerator-based constraints, exploring specific experimental signatures for these models. In the absence of flavour-breaking spurions, constraints from heavy lepton lifetime are found to dominate over most of the parameter space, with specific parts still open for future ee-collider searches. Numerical predictions are obtained using the \texttt{MARTY} framework, allowing us to consistently explore both the light- and heavy-mediator regimes. Finally, using \darkpack, we also assessed that extensions of such models could be compatible with dark matter relic density bounds.

We introduce a basis for tree-level meromorphic scattering amplitudes suitable for describing pion scattering in the large-N limit. The basis is constructed as linear combinations of Lovelace-Shapiro-like amplitudes with varying Regge slopes and intercepts. The resulting amplitudes satisfy by construction the fundamental requirements of analyticity, crossing symmetry, and Regge behavior. We analyze their behavior in specific kinematical regimes, including the high-energy fixed-angle limit. We also show that finite linear combinations of our basis elements need not violate unitarity. Nonetheless, because unitarity is not imposed by construction, we enforce it a posteriori by requiring positivity of the partial-wave decomposition. This condition can be formulated as an optimization problem and solved numerically. The solutions to this primal bootstrap problem yield meromorphic amplitudes that satisfy all the aforementioned constraints. We compare several observables with the bounds obtained from the dual positivity conditions and show that our family of amplitudes spans the full allowed parameter space. With appropriate modifications, this method can be extended to construct amplitude families for broader applications.

We report a measurement of the muon neutrino charged-current (CC) interaction cross section on tungsten using the electronic detectors of the SND@LHC experiment at the CERN Large Hadron Collider. The analysis uses proton--proton collision data at a centre-of-mass energy of $\sqrt{s} = 13.6$ TeV, corresponding to an integrated luminosity of $68.6 ~\text{fb}^{-1}$ collected during LHC Run 3 in 2022 and 2023. A total of 31 $ν_μ$ CC candidates are selected against an expected background of $5.0 \pm 1.1$ events, consistent with a signal expectation of $24^{+10}_{-9}$ events. The signal strength is measured to be $\hatμ = 1.09^{+0.72}_{-0.37}$, and the combined muon neutrino and anti-neutrino CC cross section on tungsten is determined to be $σ(ν_μ+ \barν_μ) = (37^{+24}_{-12})\times 10^{-35}~\text{cm}^2$ at a median energy of $228$ GeV. In addition, a calorimetric measurement of the hadronic energies of the neutrino candidate events is performed, making use of calibration data from dedicated test-beam campaigns.

We perform a theoretical study of the $B^+ \to D^+ D^- K^+$, $B^0 \to D^+ D^- K^0$ and $Λ_b \to D^+ D^- Λ$ reactions by looking at the production mechanisms, with special emphasis in the final state interaction of the charmed mesons, which gives rise to the $X_{c0}(3930)$, coupling strongly to $D_s\bar{D}_s$, and another state that we call $X(3700)$, coupling strongly to $D \bar D$ that qualifies as a $D \bar D$ bound state. The combined study of all these reactions shows that the final state interaction responsible for the production of these resonances is more important in the $B^+ \to D^+ D^- K^+$ reaction than in the $Λ_b \to D^+ D^- Λ$ one. We have taken this into account and shown that normalizing the $D^+ D^-$ mass distributions to the same value at the peak of the $ψ(3770)$ production, the two mass distributions are quite different in a range of 10 MeV above the $D^+ D^-$ threshold, with a value about 13 times bigger for the $B^+ \to D^+ D^- K^+$ reaction than for the $Λ_b \to D^+ D^- Λ$ one. We make a call to measure these magnitudes in the coming LHCb upgrades, which would bring great support to the existence of the bound $D^+ D^-$ state.

Contrary to naive expectations, poles used to define hadron resonances rigorously in the physical world may not evolve continuously to become degenerate in the SU(3)$_F$ and chiral limits of QCD. Instead, other shadow poles, usually ignored, may be the ones that degenerate and characterize the resonances in these limits. This feature is general, and we illustrate it first with the simple and familiar light-vector mesons, followed by the much-discussed light-scalar case. Their shadow poles and their degeneracy are found using the QCD low-energy effective theory unitarized to one loop.

We present a new version of the Sartre event generator for exclusive diffraction at small xP in the colour dipole model, for $ep$ and $e$A scattering at the EIC and the LHeC as well as ultra-peripheral $pp$, $p$A, and AA collisions at RHIC and the LHC. Sartre stores the first and second moment of the interaction amplitudes in lookup tables which are then used for efficient event generation. There are many possible combinations of processes in Sartre, with different initial state nuclear targets and different final state vector mesons or real photons. We also want to implement different versions of the dipole model: with and without non-linear saturation effects, with and without nucleon hotspot substructure. A long-standing bottleneck for simulating all possible exclusive processes has been the production of lookup tables, which take a few CPU-year for each combination, necessitating the use of computing farms. The calculation also involves integrations of rapidly fluctuating integrands, which may cause numerical glitches which takes much effort to smoothen out. In this paper we present a solution to these issues, by presenting a new numerical calculation which improves the efficiency in the table production by 3-4 orders of magnitudes. This enables us to produce lookup tables for any process that we may be interested in, in a few hours. The new calculation does also not exhibit numerical glitches. We provide novel predictions for the EIC and the LHeC using the new version of Sartre.

We study constraints on the quantum chromodynamics (QCD) axion in F-theory, a strongly coupled limit of string theory. We build models of QCD from compactifications to four dimensions on elliptic fibrations over toric threefolds $B_3$, characterized by an integer $N=h^{1,1}(B_3)$. The QCD axion mass increases with $N$, and we find that models with sufficiently high $N$ are inconsistent with observations of the neutrino burst from supernova 1987A, disfavoring large regions of the moduli space. Specifically, at least $95\%$ of models with $N \ge 8,791$ -- a regime that arguably contains the vast majority of known F-theory topologies -- have a QCD axion mass $m_{QCD}>15~meV$ and are thus constrained. This limit is independent of cosmology. We only consider weakly-curved threefolds, where $α'$ corrections are plausibly negligible.

We demonstrate that all anti-self-dual Einstein metrics with non--zero cosmological constant $Λ$ locally arise from solutions of a single second order PDE introduced by Lipstein and Nagy. We show how this equation fits into the hyper--heavenly formalism of Plebański, and establish a Lax pair. Finally we show how Plebański's second heavenly equation arises in the limit as $Λ\rightarrow 0$.

We apply Lüscher's method to the vector channel of the scattering amplitude of Pseudo-Nambu-Goldstone-Bosons (PNGBs), in the $Sp(4)$ lattice gauge theory coupled to $N_f=2$ flavors of Wilson-Dirac fundamental fermions. We generalize existing algorithms and numerical implementations of the method, to adapt them to this prominent candidate for the completion of proposed extensions of the Standard Model (SM). We present the first ab initio measurements of key properties of the vector resonances in the theory, including the coupling to the PNGBs, that are relevant to direct and indirect new physics searches, both for composite Higgs models (CHMs), as well as for strongly interacting massive particle (SIMP) realizations of dark matter. We also present a global update of the spectroscopy of the mesons in the theory, improving both the statistics and analysis systematics in respect to previous lattice measurements reported in the literature.

We provide a complete proof that Yang--Baxter bi-vector deformations are solution generating transformations in Type II supergravity. The proof does not rely on the assumption that the vielbein is invariant under Lie derivative along all Killing vectors entering the bi-vector. We analyse transformation of boundary terms and derive a simple expression for the case, when the boundary does not change its geometrical locus.

The growth of primordial density fluctuations in the early Universe leads to an inhomogeneous dark matter distribution at high redshift, where semi-annihilation processes of the form $χχ\rightarrow χ^c ν$ can occur with a sizable rate. Using a state-of-the-art model for the cosmological boost factor, we compute the resulting redshift-dependent flux of boosted dark matter particles generated by semi-annihilation, and we study the implications of the boosted component for structure formation and direct detection experiments. We find that the cosmological contribution can enhance the sensitivity of current dark matter searches by up to three orders of magnitude, and may allow future experiments to probe scattering cross sections in the femtobarn range for dark matter masses between $\sim 10$ MeV and $\sim1$ GeV.

We describe the GPU implementation of the Seeding and Matching algorithms, developed for the first level trigger of the LHCb experiment and key to reconstruct long and very displaced tracks at 40 MHz. The algorithms have been participating in the data taking during the full Run 3 of LHCb with a very high throughput, increasing the physics reach of the experiment. The Seeding is a standalone pattern recognition algorithm aiming at finding charged particle trajectories in the most forward tracker of LHCb. These trajectories are then extrapolated backward by the Matching algorithm which combines them with stubs formed from hits in the first tracker in order to form what we call Long tracks. Hits in the second tracker are then searched for to better define the trajectory and improve the track momentum resolution. This backward approach, complementary to the approach of extrapolating the stubs in the first detector to the forward tracker through the magnetic field, improves the Long track efficiency at low transverse momenta, increasing the potential of key physics decay channels.

We investigate the brane origin of supersymmetric and scale-separated AdS$_3$ flux vacua arising in a class of type IIB orientifold reductions with G$_2$-structure. First, using the effective three-dimensional supergravity together with explicit uplift formulae, we trace back the $F_{(7)}$ and $F_{(3)}$ fluxes responsible for scale separation and identify the type IIB background that is probed by the corresponding D1- and D5-branes. Second, by restoring such D1- and D5-branes, we obtain a type IIB solution that interpolates between the previous background in the asymptotic region and the scale-separated AdS$_3$ flux vacua in the near-horizon region. Third, by working directly in ten dimensions, we construct a codimension-one D1-D5-KK5 intersection whose near-horizon region realises the scale-separated AdS$_3$ flux vacua. Our results provide a higher-dimensional interpretation of these flux vacua in terms of (smeared) brane configurations.

We study entanglement measures for oscillating neutrinos in the wave-packet formalism and confirm that an oscillating neutrino exhibits genuine tripartite entanglement and can be characterized as a member of the W class of states. We also show that this feature survives in the decoherence limit, and discuss the effect of CP phases on the entanglement. These findings provide new insights into the quantum nature of neutrino oscillations and their potential role in quantum information science, especially for neutrinos propagating large distances such as the astrophysical neutrinos observed at neutrino telescopes.

Hybrid baryons, in which gluonic degrees of freedom play an explicit dynamical role, provide a key testing ground for nonperturbative quantum chromodynamics. In this work, we investigate the mass spectrum of light hybrid baryons composed of identical quarks within a phenomenological constituent framework, applied to a quark core-gluon approximation. In this approach, the hybrid baryon is described as a bound state of a color-octet three-quark core and a constituent gluon, allowing the original four-body problem to be reduced to a three-body calculation followed by an effective two-body treatment. The spectrum of the color-octet quark core is obtained by solving a semirelativistic three-quark Hamiltonian with linear confinement, Coulomb, and regularized hyperfine interactions using an oscillator basis expansion. Finite-size effects of the core are incorporated through the convolution of the effective core-gluon interaction with the spatial quark density. The resulting two-body problem, whose associated Hamiltonian has the same shape as the one of the core, is solved applying the helicity formalism and using the Lagrange mesh method. Our results predict the lightest hybrid baryons to occur at energies above $3~\mathrm{GeV}$, with negative-parity states generally lying below their positive-parity counterparts. The predicted spectra are compared with lattice QCD and QCD sum-rule calculations, showing qualitative agreement although the lowest-lying lattice QCD results are significantly lighter than the present ones. Possible extensions of the model and implications for future experimental searches are discussed.

We compute the decay width of neutrinos in a thermal medium using finite-temperature quantum field theory, focusing on non-standard decays into lighter neutrinos and a scalar or light vector boson. We derive general expressions for the thermal decay rate and show that finite-temperature effects can dramatically enhance neutrino decays when the parent and daughter states are nearly degenerate in mass. In this regime, the emitted boson is kinematically soft and undergoes strong Bose enhancement, leading to decay widths that can exceed their vacuum values by a couple of orders of magnitude. We demonstrate that this effect is largely insensitive to the Lorentz structure of the underlying interaction and arises generically from the interplay of thermal occupation factors and quasi-degenerate kinematics. Our results highlight a previously underappreciated feature of neutrino decay in thermal environments and provide a general framework applicable to a broad class of fermionic decay processes.

This paper presents a search for semi-visible Higgs boson decays in final states with a photon ($γ$) and missing transverse momentum $p_T^{miss}$. The analysis is optimised for the gluon-gluon fusion production channel and exploits a novel trigger deployed in 2023, with reduced thresholds down to 50 GeV on the photon transverse momentum ($p_T^γ$) and 70 GeV on the $p_T^{miss}$, in combination with a selection on the transverse mass of the $p_T^{miss}$-$γ$ system ($m_T$) above 80 GeV. The results are interpreted in terms of Higgs boson decays into a $γ$ and a dark photon $γ_d$ ($H \to γγ_d$). The analysis is performed using 135 fb$^{-1}$ of Run 3 data collected by ATLAS in 2023 and 2024. A boosted decision tree is used to identify events with misreconstructed primary vertices, which can affect the $p_T^{miss}$ calculation. Data-driven methods are employed to estimate the backgrounds from jets and electrons misidentified as photons, while the backgrounds with genuine photons are taken from Monte Carlo simulations, normalised to data in control regions with muons. A binned maximum-likelihood fit to the $m_T$ distribution is performed to evaluate the compatibility of the observed data with the Standard Model expectation. No significant excess is observed over the Standard Model prediction, and an observed (expected) upper limit on the branching ratio of $H \to γγ_d$ of 1.4% (1.2%) is obtained at 95% confidence level. Combining this result with previous Run 2 searches improves the observed (expected) upper limit to 0.9% (0.9%).

Using two distinct approaches, we derive a universal expression for the leading finite-volume effects of the smeared vector-vector spectral density (proportional to the smeared hadronic $R$-ratio) in a periodic cubic spatial volume of side length $L$. First, building on the results of previous work for finite-volume effects on Euclidean two-point functions, we show that the $L$ dependence is exponentially suppressed for a certain class of smearing kernels, and that the leading effects can be expressed universally in terms of the pion form factor. The same representation is then derived starting from the Lellouch-Lüscher-Meyer expression for the spectral decomposition of the correlator. The results may prove useful for controlling the $L \to \infty$ extrapolation of smeared spectral densities, in particular by defining a scaling regime in which the finite-volume effects are dominated by the leading terms in a large $L$ expansion and thus can be reliably estimated. To illustrate this point, we also present numerical estimates based on various kernels and models of particle interactions. Despite focusing on the vector channel, our derivation defines a general framework applicable to other cases as well.

We investigate a class of quintessence models in which the dark-energy scalar field interacts with sterile neutrinos responsible for neutrino mass generation through the seesaw mechanism. Radiative corrections involving sterile neutrinos induce Higgs-quintessence mixing and provide an effective portal between quintessence and Standard Model particles. We calculate the corresponding one-loop mixing amplitude and show that its structure depends on the relation between the characteristic momentum transfer and the sterile-neutrino mass. In the low-momentum regime the mixing is effectively kinetic, while at high energies it acquires a mass-mixing form. A notable result is that the overall suppression of the induced mixing is governed by the physical neutrino mass scale, leading to a predictive relation between neutrino properties and Higgs-quintessence interactions. The resulting loop-induced effects are found to be suppressed by factors comparable to those controlling the tree-level quintessence-neutrino interaction. The induced mixing generates effective couplings of quintessence to Standard Model fermions and gauge bosons, leading to modified Higgs, W and Z boson processes and opening decay channels for the quintessence field, which may prove to be cosmologically important. While the corresponding effects remain well below current experimental sensitivities in the minimal heavy-seesaw scenario, the framework establishes a direct connection between dark-energy dynamics, neutrino mass generation and Higgs-sector phenomenology and provides a basis for studying scenarios with lighter sterile neutrino states where observable effects may be enhanced.

We demonstrate that the $D_{s0}^{*}(2317)$ can be described as a bound-state pole arising from the coupling between a discrete $c\bar{s}(1^3P_0)$ state and the $DK$ continuum state in the Lee-Friedrichs model. The elementariness and compositeness of the $D_{s0}^{*}(2317)$ are determined to be about $Z:X\approx 51.1 \%:48.9\%$, indicating a nearly equal admixture of compact quark-model state and hadronic molecular components. The decay width of $D_{s0}^{*}(2317)\rightarrow D_s^{+}π^0$ is evaluated within the quark rearrangement framework. The transition $c\bar{s}(1^3P_0)\rightarrow D_sπ^0$ proceeds via the OZI-allowed $c\bar{s}(1^3P_0)-D_sη$ coupling followed by $η-π^0$ mixing, which serves as the primary source of isospin violation in this channel. The coupling between $DK$ and $D_sπ^0$ is computed in a quark rearrangement model, where the isospin violation effect originates from the mass difference between the $D^0K^+$ and $D^+K^0$ thresholds. The parameters of the scheme shares the same ones with those of the underlying potential model and only the vacuum production strength is adjusted to reproduce the physical $D_{s0}^{*}(2317)$ mass. This calculation may shed more insight to the nature of exotic $D_{s0}^{*}(2317)$ state and its isospin-breaking decay properties.

We construct light-front transverse gravitational distributions associated with the scalar trace form factor and the non-scalar gluon energy-momentum tensor (EMT) combination selected by compact-quarkonium chromoelectric scattering. The construction is performed in the Drell--Yan frame, where $t=-Δ_\perp^2$, and separates universal EMT form-factor shapes from the forward chromoelectric normalization. The resulting non-scalar distribution is normalized by the previously derived threshold strength $R_{\rm LF}^{\rm int}=N_{\rm nt}(0)/N_θ(0)\simeq 0.15$, while its transverse profile is governed by the off-forward combination of $A_g(t)$, $D_g(t)$, and $\bar C_g(t)$. We show that scalar and non-scalar responses can have different transverse localization, and that near-threshold quarkonium production probes a chromoelectric EMT projection rather than an individual form-factor slope.

We propose the One Scalar Theory (1ST), a minimalist framework where a single real singlet scalar field mediates the dynamical generation of the Higgs self-coupling and the top Yukawa coupling. Unlike generic portal models, the 1ST removes parametric freedom by locking production and decay modes to a single fundamental scale $Λ_0$, rendering the framework highly predictive with unique experimental signals. We demonstrate that the collider phenomenology is partitioned by the $2m_t$ kinematic threshold into di-Higgs and di-top resonance regimes. By recasting current ATLAS data, we set lower bounds on $Λ_0$ at the TeV scale and show that the High-Luminosity LHC will probe this scale up to $4$ TeV, providing a definitive test for the dynamical origin of the electroweak sector.

A general solution of the Balitsky Kovchegov (BK) equation well describes the partonic behavior within the saturation regime and beyond this limit. This solution can potentially define the unpolarized color dipole gluon transverse momentum dependent (TMD) distribution in the small x regime for full k perp range. In this work, we Fourier-Bessel transform this general S matrix of the BK equation and, within the leading logarithmic approximation, obtain a closed form expression of unpolarised Gluon TMD. The distribution exhibits a smooth and well behaved k perp dependence from low to high transverse momenta. Numerical evaluation for different values of x accessible at the Electron Ion Collider (EIC), shows a characteristic inversion of the x ordering at the saturation scale. The result is in close agreement with the behavior observed in the unintegrated color dipole gluon distribution computed within the Bartels Golec Biernat Kowalski (BGK) model, suggesting that this feature is a model independent signature of gluon saturation.

We present the first study of short-baseline neutrino oscillation searches using a tagged neutrino beamline, taking the proposed nuSCOPE facility at CERN as a benchmark. In this Letter we demonstrate that tagged neutrino beams, where the neutrino flavor, energy, and propagation distance are determined with exceptional event-by-event precision, provide a new experimental approach to search for non-standard neutrino oscillations. We evaluate the sensitivity to sterile-neutrino-induced oscillations in the $ν_μ$ disappearance, $ν_μ\rightarrow ν_e$ appearance, and $ν_e$ disappearance channels, demonstrating the ability to probe multiple flavor transitions and both appearance and disappearance modes within a single experiment. Our results show that tagged beams enable sensitivity improvements to mass-squared splittings spanning several orders of magnitude while substantially reducing the dependence on neutrino flux predictions that limits conventional searches. We find that nuSCOPE can probe a broad region of parameter space motivated by existing anomalies and extend coverage into previously unexplored territory, demonstrating the strong potential of tagged neutrino beams for precision oscillation physics.

We investigate thermal dilepton production from a quark-gluon plasma (QGP) with finite baryon chemical potential ($μ_{\text{B}}$) in central Pb-Pb collisions at $\sqrt{s_{\text{NN}}}=5.02~\text{TeV}$. Recent studies suggest that sizable baryon densities can be achieved at forward rapidity even at LHC energies. We incorporate finite $μ_{\text{B}}$ into a (3+1)-dimensional hydrodynamic framework and find that $μ_{\text{B}}$ exceeds 500 MeV around $η_\text{s} = 6$ during the medium evolution. Using this framework, we calculate thermal dilepton spectra over a wide rapidity range and evaluate the impact of finite $μ_{\text{B}}$ on dilepton production. A suppression of 3-4% is observed in the forward-rapidity region $5.2 < y < 7.2$ due to the reduced quark-antiquark abundance at finite baryon density. We further examine the effective temperature extracted from dilepton mass spectra in the intermediate-mass region $1.2 < M_{\ell \ell} < 2.6~\text{GeV}$ . The effective temperature remains strongly correlated with the underlying hydrodynamic temperature and retains sensitivity to the early high-temperature stage of the QGP evolution. These results demonstrate that forward-rapidity dileptons remain effective thermometers while providing sensitivity to finite baryon density at the LHC.

NOvA has reported a significance of mass ordering determination using ten years of data together with external constraints from reactor-based experiments. The JUNO collaboration is poised to provide a more precise reactor-based constraint on $|Δm^2_{32}|$. In this Letter, we explore the potential impact of this anticipated measurement on the determination of the neutrino mass ordering by NOvA. We find that $3σ$ evidence of the normal ordering is achievable over a range of plausible JUNO measurements within the next five years.

We report the first evidence for the hadronic transition $ψ(3770) \to π^0 J/ψ$ using a data sample of $20.3~\mathrm{fb}^{-1}$ collected at $\sqrt{s} = 3.773$~GeV with the BESIII detector. The $e^+ e^- \to π^{0}J/ψ$ process is observed with a statistical significance of $6.1σ$, while the significance specifically attributed to $ψ(3770) \to π^{0}J/ψ$ is $4.5σ$. We measure the dressed cross section for $e^+e^- \to π^0 J/ψ$ to be $(249 \pm 44 \pm 15)~\text{fb}$ and determine the branching fraction $\mathcal{B}(ψ(3770) \to π^0 J/ψ) = (2.08 \pm 0.36 \pm 0.21\pm 0.25) \times 10^{-5}$, where the first uncertainty is statistical, the second is systematic, and the third due to a possible interference with the $ψ(3686) \to π^{0}J/ψ$ decay. This is the first determination of this branching fraction. It lies significantly below tetraquark model predictions but aligns with calculations that incorporate meson loop effects, providing crucial insight into the isospin-violating nature of $ψ(3770)$ decays.

Axions and axion-like particles (ALPs) are protected by Peccei-Quinn (PQ) symmetries that quantum gravity is expected to break. Modeling quantum gravity by Planck-suppressed PQ-breaking operators with unsuppressed Wilson coefficients and random phases, we quantify the fine-tuning required for an acceptable strong CP phase or a given ALP mass. For the QCD axion to account for the observed dark matter abundance at $f_a \simeq 10^{11}\,\text{GeV}$, PQ-breaking operators must be absent up to mass dimension $D \gtrsim 12$. We show that the residual strong CP phase could be measurable in upcoming neutron electric dipole moment searches. For ALPs, we map the mass-decay constant plane by the degree of UV protection required, and find that parts of the parameter space targeted by future laboratory experiments are already fine-tuned at the part-per-million level or worse, or equivalently, require PQ quality to be protected up to dimension $D\gg 10$. We argue that quality, not mass alone, is the central naturalness question for the axion program.

Using a statistical model, we analyze published multiplicity distributions for identified hadrons produced in heavy ion collisions in the energy range from 2.7-200 \gev. Our analysis enables the prediction of the multiplicity distributions for multi-strangeness hadrons that have not yet been measured in lower-energy experiments. Furthermore, we obtain freeze-out parameters, including temperature, baryon chemical and strangeness potentials, the strangeness suppression factor, and the system radius, as functions of centrality and collision energy. Additionally, we computed and discussed the Skewness and its behavior at lower collision energies, highlighting the importance of the freeze-out parameters in determining the liquid-gas phase transition in nuclear matter.

In strong magnetic fields, charged quarks occupy the lowest Landau level, leading to an effective dimensional reduction of hadronic dynamics. This dimensional reduction naturally generates a hierarchy of scales, separating fast intra-meson quark dynamics from slow collective meson motion; this motivates a Born-Oppenheimer description of meson-meson systems. Our Born-Oppenheimer analysis shows that the infrared behavior is controlled by the interplay between dimensional reduction and the structure of the meson-meson interaction, leading to three distinct regimes: scattering-dominated, molecular, and compact multiquark states. Charged pseudoscalar mesons such as $π_+$ and $K_+$ provide a particularly interesting realization of this framework, as their lattice spectra at large magnetic fields suggest the emergence of loosely bound states near the boundary between scattering and molecular regimes. Our results suggest that strong magnetic fields provide a useful laboratory for exploring the emergence and classification of hadronic bound states.

In the context of neutrino oscillation measurements from the T2K experiment, the off-axis near detector ND280 plays a crucial role in constraining the incoming neutrino flux and neutrino-nucleus interaction cross sections. The result is a robust control over systematic uncertainties in the fit of neutrino oscillation parameters to the data at the T2K far detector, Super-Kamiokande. This paper details the methodology and results of these constraints in the context of the latest neutrino oscillation analysis from T2K. It describes how a new neutrino cross-section model and refined flux prediction are parameterized and fit to data in new ND280 event selections. Additionally, this work reports the results of extensive robustness studies, including fits with alternative interaction models, consistency checks against publicly available cross-section measurements, and \textit{p}-value evaluations, to demonstrate the reliability and robustness of our methodology. Finally, we present a sensitivity study demonstrating that the upgraded ND280, with improved acceptance and a lower hadron threshold, may enhance future constraints and further reduce systematic uncertainties in oscillation measurements.

Future neutrino experiments, in particular DUNE, are expected to probe signals of new physics. Those can be conveniently parametrized in terms of Wilson coefficients in the LEFT, with direct connection to non-standard interactions at production, propagation and detection in the QFT formalism. Assuming that the new degrees of freedom lie above the electroweak scale, we investigate if a DUNE-motivated signal, described by a non-vanishing SMEFT Wilson coefficient, can be realized in UV completions containing extra isosinglet and/or isodoublet fields, once present flavour constraints are taken into account. To this end, we propose a pipeline which can, in principle, be applicable to any BSM signal parametrized by a given set of SMEFT Wilson coefficients. As a concrete example, we focus on the lepton-flavour-violating semileptonic coefficient $(C^{(1)}_{\ell q,1311})$. We find that the largest viable value of the target Wilson coefficient remains around ($10^{-2}\;{\rm TeV}^{-2}$), which is almost one order of magnitude below the DUNE benchmark.

We investigate the chemical-potential dependence of hadron spectra in two-color QCD using first-principles lattice simulations. We compute two-point correlation functions for all allowed hadronic operators by newly including the contributions from disconnected diagrams, and extract the corresponding effective masses. In the meson sector, the mass hierarchy in the hadronic phase (normal vacuum) is found to be $m_π\lesssim m_η < m_σ\mathrm{(noisy)} < m_ρ\sim m_ω\ll m_{a_1}$, which is similar to that in three-color QCD. In the superfluid phase, this hierarchy is modified, and with increasing density it changes to $m_σ\mathrm{(noisy)} < m_{a_1} < m_ρ< m_π\sim m_η \mathrm{(noisy)} \ll m_ω \mathrm{(noisy)}$. In the diquark sector, the ordering remains as $m_{NG} \lesssim m_{I=0, S} < m_{I=1, AV} < m_{I=0, PS} \lesssim m_{I=0, V}$ in both phases, and the Nambu--Goldstone mode associated with spontaneous breaking of $U(1)_B$ is confirmed to be nearly massless. Furthermore, by comparing correlators for chiral partners, we find indications of chiral symmetry restoration at high density.

Motivated by the recent interest in multipoint Padé approximants in the physics community, we discuss Chebyshev rational functions and show how they give rise to multipoint Padé approximants in exactly the same way that Chebyshev polynomials produce Padé approximants. We present recurrence relations for Chebyshev rational functions, as well as the underlying continued fraction of Thiele type (also known as $R_{II}$ type). Finally, we provide numerical evidence illustrating the effects of noise on this interpolation scheme and show that a phenomenon similar to that recently observed by Costin, Dunne, and Meynig for Padé approximants also occurs in the multipoint setting.

New binning schemes for the Dalitz plots of $D \rightarrow K^0_{\mathrm S}π^+π^-$ and $D \rightarrow K^0_{\mathrm S} K^+ K^-$ decays are presented. These are determined using a new figure of merit for optimisation that better represents the sensitivity to $γ$ than previous metrics. Further augmentation comes from including consideration of degeneracy with other physics parameters determined simultaneously and a more accurate treatment of relevant backgrounds. The overall expected improvement in the precision of $γ$ measurements using $B^\pm\to DK^\pm$ decays, followed by $D \rightarrow K^0_{\mathrm S}π^+π^-$, is estimated at around 5$\%$. In addition, the first dedicated optimisation is performed for the observables relevant to charm mixing, $x_{CP}$ and $Δx$, in $D \rightarrow K^0_{\mathrm S}π^+π^-$ decays. This procedure accounts for both statistical sensitivity and the migration of events between bins due to detector resolution. The resulting binning scheme leads to an estimated gain of approximately 20$\%$ in statistical sensitivity, while maintaining the bias due to bin migration at a level comparable to the currently used scheme.

We employ a model-independent Effective Field Theory (EFT) to analyze the possibility of a strong First-Order Phase Transition (FOPT) in extensions Beyond the Standard Model (BSM). We find that sizable deviations in the Higgs cubic and quartic interactions that are still allowed experimentally could lead to a strong FOPT, whereas the Higgs interactions to the top quark yield a weak FOPT. We also study the Gravitational Wave (GW) power spectra corresponding to the strong FOPT and find that they could be detectable in future experiments. In particular, we find that deformations of the Higgs quartic coupling have the dominant impact on the FOPT, with a GW signal that could be probed by a number of future experiments, such as LISA, BBO, and DECIGO. We also study the magnetic field produced by the corresponding FOPT and find that it could explain the primordial magnetic field puzzle. We find that for the size of deformations that could induce a strong FOPT, a scale of NP can be as low as $\sim 4\text{--}5~\text{TeV}$ for deformations in the Higgs cubic coupling, and $\sim 9\text{--}11~\text{TeV}$ for deformations in the Higgs quartic coupling. This highlights the synergy between collider searches and GW experiments in probing the Higgs couplings, specifically the Higgs quartic coupling.

We investigate the interactions of the recently confirmed exotic hadron $G(3900)$ with the hadronic medium composed of light mesons produced in heavy-ion collisions. Using an effective Lagrangian approach, we compute both vacuum and thermally averaged cross sections for processes such as $G(3900) + π\to D^{(\ast)} + \bar{D}^{(\ast)}$ and their inverse reactions. Following recent proposals, the $G(3900)$ is interpreted as a $P$-wave $D\bar{D}^*/D^*\bar{D}$ molecular state. The resulting thermally averaged cross sections are employed in a rate equation to determine the time evolution of the $G(3900)$ multiplicity, with initial conditions provided by statistical and coalescence hadronization models. We compute the number of $G(3900)$'s produced in central Pb-Pb collisions at $\sqrt{s_{NN}} = 5.02$ TeV and compare it with the number of produced $Z_c(3900)$'s . We obtain a considerably smaller final yield of $G(3900)$ compared to $Z_c(3900)$.

Quantum Chromodynamics (QCD) provides a remarkable realization of how quantum effects reshape the symmetries of a classical field theory. The axial anomaly links chirality to gauge-field topology and underlies the resolution of the $U(1)_A$ problem, while the trace anomaly generates the intrinsic QCD scale through dimensional transmutation and accounts for most of the mass of hadrons and hence visible matter. Together, these quantum effects reveal the central role of gluonic dynamics and vacuum structure in strong-interaction physics. In this review, we discuss the theoretical foundations and physical consequences of anomalous symmetry breaking in QCD. We examine anomalous Ward identities, the topological structure of gauge fields, instantons, topological susceptibility, and the realization of chiral and scale symmetries in the QCD vacuum. We review their role in the generation of hadron masses, the $η'$ mass, and the resolution of the $U(1)_A$ problem, and discuss their manifestations in polarized deep inelastic scattering, nucleon spin structure, and modern studies of hadron structure. Particular emphasis is placed on recent developments connecting vacuum topology, the flavor-singlet axial charge, topological screening, and the proton spin problem. Our aim is to provide a unified perspective on how quantum anomalies connect vacuum structure, hadron properties, and partonic observables, bridging nonperturbative dynamics and perturbative QCD.

We study holographic entanglement entropy in Coulomb-branch solutions describing spherical shells of D$p$-branes. The corresponding throat geometries contain a flat-space bubble in the infrared region, providing a concrete top-down framework for exploring holographic entanglement of flat space. We find that the flat-space region is associated with a reduction of entanglement and of the effective infrared degrees of freedom in the dual boundary state relative to the standard vacuum. We also examine internal RT surfaces and holographic complexity, and show that they exhibit similar qualitative behavior. Finally, we comment on the broader implications of our results for flat space holography.

Building on two key ingredients, namely the well established gauge-theoretic formulation of gravity and the observation that the tangent space of a curved manifold need not have the same dimension as the manifold, we discuss how all known fundamental interactions can be accommodated within a single unified framework. The unification is realised by enlarging the tangent group of the four-dimensional spacetime manifold to $SO(2,16)$, a choice that simultaneously encompasses both the gauge group underlying the gravitational sector and the $SO(10)$ Grand Unified Theory for the internal interactions. The gravitational theories entering this construction are Conformal Gravity and Fuzzy (Noncommutative) Gravity, each formulated in gauge-theoretic terms.

We study the renormalization-group flow of higher-dimensional fermionic interactions $(ψ^\dagger ψ)^{2n}$ in the presence of a Fermi surface. We show that the growth of the BCS four-fermion coupling induces a drag mechanism whereby higher-order fermionic couplings are driven to strong coupling in the infrared. We derive the corresponding beta functions and show that the drag effect applies generically to the whole tower of fermionic operators. Remarkably, although all couplings are driven toward the same strong-coupling scale, the renormalization-group flow preserves a hierarchy in which higher-dimensional operators remain parametrically suppressed relative to the BCS interaction. We then investigate this mechanism in a $2+1$-dimensional non-Fermi liquid coupled to a critical boson. While higher-order fermionic interactions are similarly enhanced along the flow, we show that they do not destabilize the IR stable non-Fermi-liquid fixed point, present for sufficiently large $N$. We briefly discuss possible implications for multicomponent superconductors and other strongly correlated metallic systems.

The primordial black hole (PBH) mass function today develops a low-mass evaporation tail whose shape is universal with respect to the initial PBH mass distribution. This universality is fixed by the continuity equation and the Hawking mass-loss rate, but the tail is distorted if additional particle degrees of freedom participate in Hawking evaporation. Since this tail controls high-energy PBH photon emission, such distortions leave characteristic features in the gamma-ray spectrum. We show that these features provide a robust probe of new degrees of freedom, even for subdominant PBHs, within reach of future experiments and independent of visible-sector couplings or relic abundance.

We investigate the small circle, or high temperature, limit of the supersymmetric index identities for the three-dimensional abelian mirror symmetry of SQED. There exist two possible limits, depending on how the parameters of the theory are scaled with the radius of the circle. In both cases the result is qualitatively similar. One side always reduces to the sphere partition function of a two-dimensional $\mathcal{N}=(2,2)$ gauged linear sigma model (GLSM). The opposite side has a two-fold interpretation, either as the sphere partition function of the Landau--Ginzburg (LG) model that is Hori--Vafa dual to the GLSM, or as a Coulomb gas integral for a correlation function of Liouville or Toda CFT. This approach thus provides a systematic way to generate integral identities between partition functions of GLSMs on the one hand, and partition functions of LG models or CFT Coulomb gas integrals on the other. The latter perspective finds useful applications in the recently proposed 2d/2d correspondence, which relates sphere partition functions of unitary 2d $\mathcal{N}=(2,2)$ theories and correlation functions of non-unitary 2d CFTs that both descend from compactifications of a unitary 4d $\mathcal{N}=2$ SCFT. We present an example based on the $(A_{k-1},A_{N-1})$ Argyres--Douglas theories, where the CFT is a non-unitary minimal model. We also give a purely two-dimensional derivation of the identities obtained in the small circle limit which is inspired by the Kapustin--Strassler piecewise derivation of 3d abelian mirror symmetry.

We derive spectral identities involving the Laplace spectrum and the integrals of products of eigenfunctions on flat tori and orbifolds. Using semidefinite programming, we derive upper bounds on sums of squares of triple products from these spectral identities, following the approach of the conformal bootstrap. Physically, these upper bounds give constraints on sums of squares of cubic coupling constants in toroidal compactifications of higher-dimensional field theories. For Euclidean lattices, one of the bootstrap bounds leads to an upper bound on the mean square number of minimal vectors that are minimal distance from each minimal vector, divided by the kissing number of the lattice. By finding exact functionals for the semidefinite programming problems, we prove that this bound is saturated in 2, 4, 8, and 24 dimensions by the hexagonal lattice, the $D_4$ lattice, the $E_8$ lattice, and the Leech lattice, respectively.

We study dynamical mass generation of a fermion with and without a bare mass by coupling with a massive vector field in two-dimensional space-time. To estimate a non-perturbative effect on the fermion mass, we employ the Schwinger-Dyson equations in the lowest-ladder approximation, which are solved by an approximated analytical method and also by a numerical method. We define a purely dynamical mass as a remnant after subtracting the bare mass from a total dynamical mass. We clarify dependence of the purely dynamical mass on the bare mass of the fermion in various region of a coupling constant. Especially we find that the purely dynamical masses growing from the different bare masses coincide with each other at a specific value of the coupling constant where a kind of a duality relation on the bare masses is satisfied.

Using a direct scan of 26 energy points near the $J/ψ$ resonance, we perform the first measurement of the relative phase $φ_{γ, 3g}$ between the strong and electromagnetic amplitudes in $J/ψ\to\bar K^0 K^*(892)^0+c.c.$ and $J/ψ\to K^+ K^*(892)^-+c.c.$. Unexpectedly, the phase in the neutral channel is found to be $\sim 150^\circ$, deviating from orthogonality ($90^\circ$) by 4.2$σ$ and from a relative real amplitude (0$^\circ$ or 180$^\circ$) by 10.0$σ$ or 1.6$σ$, respectively. In contrast, the charged channel phase is consistent with $\sim 180^\circ$ within 1$σ$, exhibiting model-dependent behavior. The corresponding branching fractions are consistent with the world averages but achieve better than twofold improvement in precision. The ratios between the branching fractions of $J/ψ\to\bar K^0 K^*(892)^0+c.c.$ and $J/ψ\to K^+ K^*(892)^-+c.c.$ are also measured. After subtracting the electromagnetic contribution, the corresponding strong amplitude ratios obey isospin symmetry within $1.8σ$. A search for direct $CP$ violation yields asymmetries consistent with zero.

Using $e^+e^-$ annihilation data collected at 26 center-of-mass energy points between 3000.00 and $3119.88~\text{MeV}$ with the BESIII detector, corresponding to a total integrated luminosity of about $440.7~\text{pb}^{-1}$, we study the cross section lineshape of $e^+e^-\to K_S^0 K^+π^-+c.c.$. The relative phases and magnitudes between $J/ψ$ strong and electromagnetic decay amplitudes are measured to be $(123.7\pm5.3)^\circ;4.31\pm0.22$ or $(-123.1\pm5.2)^\circ;4.38\pm0.22$, with corresponding branching fractions $\mathcal{B}(J/ψ\to K_S^0 K^+π^-+c.c.)=(5.17\pm0.20)$ or $(5.36\pm0.20)\times10^{-3}$. Based on a partial wave analysis, the cross sections of $e^+e^-\to\bar K^0 K^*(892)^0+c.c.$ and $e^+e^-\to K^+ K^*(892)^-+c.c.$ are obtained. For these subprocesses, the relative phases and magnitudes are determined as $(155.2\pm15.5)^{\circ};3.67\pm0.27$ or $(-154.1\pm15.5)^{\circ};3.71\pm0.25$ and $(180.1\pm31.8)^{\circ};25.06\pm2.51$, respectively. The large relative phases deviate from the orthogonality relation expected from experiment and from the assumption of purely real amplitudes by more than $3σ$. The measured branching fractions $\mathcal{B}(J/ψ\to\bar K^0 K^*(892)^0)+c.c.=(4.18\pm0.18)$ or $(4.31\pm0.19)\times10^{-3}$, $\mathcal{B}(J/ψ\to K^+ K^*(892)^-+c.c.)=(7.09\pm0.28)\times10^{-3}$ are all consistent with the world average values, but achieve better than a twofold improvement in precision. The ratios between the branching fractions of $J/ψ\to\bar K^0 K^*(892)^0+c.c.$ and $J/ψ\to\bar K^+ K^*(892)^-+c.c.$ are $\mathcal{R}_{K^*\bar{K}}=0.589\pm0.012$ or $0.612\pm0.013$. After subtracting the electromagnetic contribution, the corresponding strong amplitude ratios are $\mathcal{R}^{3g}_{K^*\bar{K}}=0.884\pm0.050$ or $0.909\pm0.044$, which deviate $2.3$ or $2.1σ$ from the unity. No evidence for direct $CP$ violation is observed.

The 21-cm signal probes the intergalactic medium during the Epoch of Reionization (EoR) but is overwhelmed by astrophysical foregrounds orders of magnitude stronger than the cosmological signal. We evaluate two mitigation strategies: (i) Foreground Avoidance, restricting analysis to the EoR window in Fourier space, and (ii) Foreground Removal via Gaussian Process Regression, which exploits spectral smoothness to statistically separate contaminants and reclaim modes within the contaminated wedge. Both introduce systematic biases of up to $\approx$1$σ$ in astrophysical parameters such as the minimum star-forming halo mass and ionising escape fraction, with avoidance posteriors consistently broader than removal owing to its restricted visibility coverage. The global reionization history is recovered within the 95% credible interval, though the neutral fraction at late reionization epochs shows a persistent bias reflecting the difficulty of its inference from the power spectrum alone. Multi-redshift inference is susceptible to contamination from poorly mitigated bins. Excluding such bins significantly reduces parameter biases, but identifying them requires independent quality metrics. When restricted to identical length scales, both strategies recover similar power spectra, yielding posteriors in similar regions.

Understanding the sources and evolution of cosmic reionization remains a central challenge in astrophysics, with the escape of ionizing Lyman-continuum (LyC) photons from early galaxies representing a major uncertainty. In this work, we use more than 35,000 galaxy realisations from the THESAN-ZOOM cosmological radiation-hydrodynamic simulations to identify indirect diagnostics of the LyC photon escape fraction ($f_\mathrm{esc}$) and the LyC photon escape rate ($\dot{N}_\mathrm{ion,esc}$) across the redshift range $z=3-16$. We train random forest regression models using these diagnostics to predict both quantities. We present four models: two trained with the full set of simulation-derived indicators to predict $f_\mathrm{esc}$ and $\dot{N}_\mathrm{ion,esc}$, and two restricted to observables accessible to JWST photometric surveys. We find the 10-to-100$\,$Myr star-formation rate ratio ($\mathrm{SFR}_{10} / \mathrm{SFR}_{100}$) and the gas-to-stellar mass ratio ($M_\mathrm{gas} / M_*$) to be the strongest diagnostics of $f_\mathrm{esc}$, suggesting a strong relationship between ionizing photon escape and gas clearing through bursty star formation. In contrast, rest-frame UV ($1500 \, Å$) absolute magnitude ($M_\mathrm{UV}$) dominates $\dot{N}_\mathrm{ion,esc}$ prediction. Motivated by the strong predictive power of $M_\mathrm{UV}$, we combine observed UV luminosity functions with derived $\dot{N}_\mathrm{ion,esc} - M_\mathrm{UV}$ relations to construct histories of reionization. These are consistent with observational constraints, avoiding the recently reported crisis in the ionizing photon budget. Our analysis suggests that the bulk of reionization occurred rapidly after $z \approx 8$, driven by UV-bright galaxies, with the $M_\mathrm{UV} < -17$ populations providing the dominant contribution.

Aims. The increasing discovery of high-redshift AGNs in recent years imposes more stringent requirements on spectral analysis tools for deriving the properties of AGNs and their host galaxies from emission-line diagnostics. To address this need, we develop a new module for the popular SED-fitting tool Code Investigating GALaxy Emission (CIGALE), the [nebular_AGN] module, which enables the efficient and flexible simulation and fitting of emission lines originating from the broad-line regions (BLRs) and narrow-line regions (NLRs) of AGNs, and allows the estimation of the physical properties of these regions. Methods. We use the spectral synthesis code Cloudy to construct the database for the new module. Based on the X-ray and accretion disk continua implemented in CIGALE, we generate the incident radiation fields of the models. We then adopt the AGN geometry and dust settings implemented in CIGALE to define a flexible set of physical parameters for the gas clouds, thereby producing a comprehensive database for the [nebular_AGN] module. Results. We benchmark the [nebular_AGN] module using a quasar composite spectrum, an empirical metallicity calibration, and observational data from X-ray-selected AGNs. Our module can approximately reproduce the majority of quasar emission-line profiles, cover the key emission-line ratios observed in AGN samples, and provide an assessment of their physical properties. For specific combinations of parameters, the metallicity derived by our module is consistent with the empirical formula. We further compare our models with other photoionization models used to simulate AGN NLR emission, and perform a line-sensitivity study to identify the most effective diagnostic lines for each parameter in our module. Finally, we confirm that the dust attenuation law plays an important role in SED fitting.

The thermal Sunyaev--Zel'dovich (tSZ) power spectrum is a sensitive probe of cosmology and cluster astrophysics, but its statistics are non-Gaussian because the signal receives a significant contribution from rare, massive, low-redshift galaxy clusters. As a result, a Gaussian likelihood fails to describe the statistics of its power spectrum on large scales. We use simulation-based inference (SBI) to test the accuracy of the standard Gaussian power-spectrum likelihood for a \textit{Planck}-like tSZ analysis. Using halo-based simulations of full-sky Compton-$y$ maps, we train neural posterior and likelihood estimators and compare the resulting constraints with those from a Gaussian likelihood assumption. Using only multipoles $\ell < 1000$, we find that the Gaussian likelihood assumption gives unbiased cosmological constraints, while the SBI-based inference shows a mild broadening of the posterior distributions for the amplitudes of residual foregrounds. This suggests that the Gaussian likelihood assumption is sufficiently accurate for cosmological inference for a \textit{Planck}-like tSZ analysis, while SBI provides a useful validation tool to model non-Gaussian likelihoods beyond analytic approximations.

Recent DAMPE measurements of individual cosmic-ray components, including carbon, oxygen, and iron, reveal distinct spectral softenings below the knee. The energy spectrum, mass composition, and anisotropy together provide key probes of cosmic-ray origin and propagation. By incorporating the individual elemental spectra reported by DAMPE, we derive a more complete $\langle \ln A\rangle$ distribution, which smoothly connects to the higher-energy $\langle \ln A\rangle$ measurements from LHAASO and exhibits a pronounced bump-like feature. This feature indicates a transition from the conventional Galactic cosmic ray source population to a nearby-source-dominated regime. We show that a spatially dependent propagation model with a nearby-source contribution can consistently reproduce the observed spectra, mass composition, and anisotropy. This suggests a unified picture in which Galactic cosmic rays below the knee arise from multiple source populations jointly constrained by these observables. Leveraging the precise component-resolved spectra from DAMPE, we further predict the transition energies in the anisotropy phase and amplitude for different mass components. Future component-resolved anisotropy measurements by LHAASO will provide a crucial test of this scenario.

Pulse profile modelling is a relativistic ray-tracing technique used to infer neutron star mass, radius, and surface hotspot properties from X-ray pulsations. Pulse profile modelling has been widely applied to rotation-powered millisecond pulsars, where the local environment is relatively empty. Application to accreting millisecond pulsars is complicated by the geometry of the local accretion flow, including disc occultation of surface emission. In this work, we extend an established pulse profile modelling code, X-PSI, to incorporate accretion disc occultation in accreting millisecond pulsar pulse profile modelling. We quantify how disc occultation depends on system geometry and evaluate its impact on parameter inference. We find that disc occultation is primarily governed by the viewing inclination and can significantly reshape pulse profiles at moderate to high inclinations. Using synthetic Neutron Star Interior Composition Explorer datasets, we investigate parameter recovery for two representative hotspot configurations. For hotspots close to the rotational poles, statistically acceptable fits can yield posteriors that deviate noticeably from the true parameters. In contrast, in a case with hotspots located closer to the equator we find more reliable parameter recovery. We further find that neglecting disc occultation can introduce spurious posterior modes with comparable statistical support, potentially affecting the interpretation of inferred neutron star parameters, suggesting that this effect should be included in accreting millisecond pulsar pulse profile modelling.

The subarcsecond angular resolution and stable background of JWST has given us the first high-fidelity images of the arcsecond-scale environment around Active Galactic Nuclei (AGNs) in the nearby Universe. With mid-infrared (MIR) surface brightness sensitivities that are much deeper than the best ground-based instruments, the Mid-InfraRed Instrument imager (MIRIM) now allows us to understand the structure and thermal properties of dust using information over wavelengths of $5$-$25$ $μ$m, almost all of the MIR range. We present a Cycle 1 JWST MIRIM survey of Seyfert galaxies with the express aim of characterising AGN-heated dust in the central few 100 pcs, and searching for signatures of dust-laden nuclear outflows. This paper outlines the motivation behind the programme, the data reduction and analysis techniques used to isolate the nuclear and extended emission, and a comparison of the observed MIR structures with those seen in other phases (stars, ionised and molecular gas, absorbing dust). In concert with earlier studies that used these data, we conclude that resolved AGN-heated dust is widespread in the Seyfert population, extending out to a few hundred pcs from the nucleus and often displaying a higher surface-brightness compared to the more widespread star-forming dusty circumnuclear disk. Even after accounting for contamination from emission lines in the MIRI filters, we find strong spatial correlations between MIR dust emission and the AGN-ionised gas in the narrow-line region (NLR).

We analyze the mass-ratio and effective-spin ($χ_{\rm eff}$) distributions of binary black hole mergers in the latest gravitational-wave catalog, GWTC-5, as a function of primary mass. Using hierarchical Bayesian inference with flexible Gaussian-process population models, we identify four distinct mass regions separated by sharp transitions in both mass-ratio and spin properties. Below $\sim15~M_{\odot}$, the population strongly favors equal-mass binaries and exhibits a narrow $χ_{\rm eff}$ distribution peaked at positive values. In the range $18$-$30\,M_{\odot}$, the mass-ratio distribution becomes substantially flatter, while the $χ_{\rm eff}$ distribution broadens, shifts to a peak consistent with zero, and shows tentative--but not statistically required--evidence for positive skewness. The region associated with the feature near $\simeq35~M_{\odot}$ returns to a narrow $χ_{\rm eff}$ distribution consistent with symmetry at zero and strongly favors equal-mass binaries. Above $\simeq 45~M_{\odot}$, both the mass-ratio and $χ_{\rm eff}$ distributions broaden significantly. The inferred support of the spin distribution converges toward the range expected for binaries containing remnants of previous black hole mergers, making the highest-mass region fully consistent with a star cluster population of hierarchical mergers. The close correspondence between transitions in mass ratio and effective spin suggests that different primary-mass ranges trace distinct formation channels, with isolated binary or triple evolution likely dominating the lower-mass population and dynamical assembly becoming increasingly important at higher masses.

We present a morphokinematic analysis based on high-resolution long-slit echelle spectroscopy of the \nii$\lambda6583$ line and narrowband imaging. Position-velocity diagrams reveal asymmetric expansion and localized kinematic features. We derive a systemic velocity of $V_{\rm sys}^{\rm LSR} = -25\pm1$\kms\ ($V_{\rm sys}^{\rm HEL} = -34 \pm 1$\kms) and a main shell expansion velocity of $V_{\rm exp} = 22 \pm 1$\kms. Three-dimensional modeling indicates an ellipsoidal main body surrounded by a thin shell, two ear-like protrusions, and additional small-scale structures. The corresponding kinematic ages are $3\,600 \pm 700$ yr for the ellipsoid and ring, and $7\,500 \pm 1\,000$ yr and $8\,800 \pm 1\,500$ yr for the two opposite ear-like protrusions, respectively, indicating that these outer structures predate the main nebular envelope. The kinematic asymmetry and enhanced emission regions suggest evolution within a non-uniform ambient medium. At the same time, the presence of collimated ear-like structures is consistent with shaping influenced by binary interaction, where earlier outflows preceded the ejection of the dense shell. NGC\,6563 therefore appears to be a dynamically evolved system shaped by the combined effects of episodic mass ejection and environmental interaction.

In the late 20th century, the Hubble Space Telescope (HST) revolutionized astronomy, showing the Universe with a detail never seen before in the ultraviolet (UV), optical and infrared (IR) bands. In the early 2020s, the James Webb Space Telescope started a similar revolution in the IR. The launch of Roman in late 2026 challenges the reign of HST in the optical band, but even after Roman's launch, HST will remain as the only telescope capable of high-quality imaging in the UV band. In the optical bands, HST provides superior resolution than Roman for point sources. Although equipped with more sensitive MOS, Roman's sensors have a pixel size about 3 times larger than HST's CCDs, hence undersampling the point-spread-function, and resulting in a worse spatial resolution. The UV-capable and higher-resolution of HST in the UV and optical band, makes HST the best instrument for specific science cases. This paper responds to the "Building a Roadmap for Hubble science into the 2030s" call and focuses on science with lensed hot stars at $z>0.5$ in the UV and optical bands, exploiting the features that makes HST the best instrument in the UV/optical until the launch of the Habitable World Observatory in the 2040s.

We conduct two three-dimensional hydrodynamic simulations of the jittering-jets explosion mechanism (JJEM) of core-collapse supernova (CCSN), launching three pairs of inclined opposite jets into the core of an enveloped-stripped stellar model, and reproduce some morphological features of observed CCSN remnants (CCSNRs) that a single pair of jets or instabilities alone cannot reproduce. We launch the three pairs of jets within about a second, and follow the ejecta for more than 10 seconds until after shock breakout. Our main findings are: (1) Although the jets are choked deep inside the star, they manage to form a pronounced multipolar (point-symmetric) morphology. (2) Instabilities and vortices resulting from the jet-star interaction form small clumps and narrow filaments, some of which form point-symmetric morphology, resembling some observed CCSNRs. (3) The most energetic jet of one simulation forms a large low-density blowout ahead of the ejecta, with filaments dragging behind it, resembling the blowout of the Cygnus Loop. (4) The inner ejecta presents two symmetry axes along two of the three jet axes: one of a pair of rings and one of a pair of nozzles, resembling the structure of the point-symmetric SNR J0450.4-7050. (5) The three pairs of jets compress two dense blocks between their axes. The blocks exhibit a Doppler-shift bipolar outflow highly inclined to the morphological axes along the jet axes. The inclined Doppler-bipolar outflow and morphology axis resembles the CCSNRe W49B and SNR G292.0+1.8. Our study supports the claim that the JJEM is the primary explosion mechanism of CCSNe.

We report multiwavelength observations of EP240305a, an uncatalogued X-ray transient detected by the Einstein Probe on March 5, 2024. The source exhibits distinct characteristics across the X-ray, optical, near-infrared, and radio bands. The soft X-ray observations show two significant flares lasting ~100-250 s, accompanied by rapid flux decay in a few days, and the optical and near-infrared data reveal a faint, candidate counterpart. In contrast, the radio observations expose a long-term spectral evolution from a self-absorbed to an optically thin state within two months, implying discrete jet ejection. We compare EP240305a with known classes of X-ray transients and find that it is unlikely to be associated with long-timescale transients such as jetted tidal disruption events or X-ray binaries. Its properties also disfavor a short-timescale stellar flare origin. Although the absence of optical spectroscopy prevents a redshift determination, the source exhibits properties similar to those of gamma-ray-dark gamma-ray burst-like transients, which may be associated with relativistic jets viewed off-axis or with choked jets. The discovery of EP240305a, along with other uncataloged transients detected by the Einstein Probe, underscores the scientific potential of highly sensitive X-ray survey telescopes and rapid-response multiwavelength follow-up observations in exploring the nature of atypical astronomical transients.

We present a simple and robust framework to assess the information content of the power spectrum of a marked density field within the Effective Field Theory of Large Scale Structure, providing a physically motivated regime of validity of perturbation theory for such fields. We optimize the choice of mark to maximize the constraining power on cosmological parameters, focusing on the resulting reduction of parameter uncertainties. After marginalizing over the small-scale counterterms, we find improvements in the parameter error bars by a factor of approximately $1.30$-$2.3$, $1.3$-$8.3$, and $1.2$-$2.5$ for $Ω_m$, $σ_8$, and $h$, respectively, depending on redshift and smoothing scale. The optimized mark exhibits a simple functional form, for which we provide a physical interpretation in terms of its response to the underlying density field. These results demonstrate the potential of marked statistics as a tool to extract additional cosmological information beyond standard two-point analyses and open a pathway for further exploration of nonlinear transformations of the density field.

We determine the dark matter (DM) distribution in galaxies by jointly modelling stellar kinematics from IFU observations and the gaseous kinematics encoded in a single integrated HI spectrum. The stellar kinematics are described by a triaxial orbit-superposition Schwarzschild model, while the HI gas is described by an idealised disc model; both are governed by the same gravitational potential. The potential comprises the stellar mass, a generalised NFW DM halo, and a central black hole. We validate the method on 58 simulated galaxies generated from the TNG50 cosmological simulation. For each galaxy, we create two versions of mock data with azimuthal angles viewed side-on and end-on thus 116 mock observations in total. Our model recovers the total mass, stellar mass, and DM mass profiles within the data range; the median DM mass of the 58 simulated galaxies is recovered with a relative systematic bias smaller than 20% across all radii from 2--20 kpc. The statistical uncertainties on the DM masses within 5 kpc remain similar to those found with the model constrained by IFU data only. In contrast, the relative uncertainty on the DM mass in the outer regions decreases once the HI spectrum is included; at 20 kpc, it drops markedly, from about 85% to roughly 30%. The DM density slope defined explicitly in the gNFW model is systematically underestimated and thus does not yield a reliable quantity from observations using our approach. Instead, we introduce density slopes evaluated between 2 and 20 kpc, which are statistically well recovered for both the total mass and the DM mass. We demonstrate the reliability of this method in uncovering DM distribution and emphasize its promise for application to large samples of observed galaxies.

We investigate the nature of the unassociated sources detected by the Fermi-LAT close (|b|<10°) to the Galactic plane, representing 16% of all sources in the 4FGL-DR4 catalog. The bulk of these sources (referred to as soft Galactic unassociated sources, SGUs) exhibit properties not found in known classes of gamma-ray emitters, as confirmed by a machine-learning classification approach. In particular, these properties include a steep, curved spectrum peaking below 1 GeV and a specific Galactic-latitude distribution with both a narrow and a broad component (dubbed the spike and the shoulder, respectively). Some source clusters are highlighted. New plausible source classes are explored, but only star-forming regions are found to account for a significant fraction (at most 10%) of the unassociated population. A thorough search for counterparts to the 175 brightest sources brings out a number of plausible counterparts but does not reveal clues about the nature of the whole population. We investigate the possibility that SGUs originate from mismodeled clumps of diffuse emission. Using Monte Carlo simulations, the SGU spectra can be reproduced in this scenario under an ad hoc condition concerning the clump spatial extension. The possible connection between the SGUs and gas not accounted for by the 12CO tracer is explored using the 13CO MOPRA data but leads to inconclusive results. The origin of SGUs being related to diffuse emission remains plausible. However, a scenario whereby SGUs represent a new class of gamma-ray emitters cannot be fully excluded.

As the LIGO-Virgo-KAGRA collaboration's (LVK's) gravitational-wave transient catalog grows, we are learning a wealth of information from the population properties of binary black hole mergers. Events in the catalog are represented with posterior samples describing the astrophysical parameters for each event. Population studies combine these samples to measure the distribution of astrophysical parameters such as black hole masses and spins. However, the posterior-sample representation of each event is only approximate. We demonstrate that when $\gtrsim 300$ events are combined, the numerical error can become large enough that the resulting population inference is unreliable. We consider two solutions. In the short term, we show that nested samples (already produced by LVK analyses) can be used to more accurately describe each event in population studies. But this will only grant a temporary reprieve until the nested-sample representation becomes inadequate. In the longer term, we propose to represent each event with a normalizing flow. Each normalizing flow can be used to generate an arbitrarily large number of new posterior samples in order to represent each event with sufficient accuracy.

Warped galactic discs are common, yet their detection remains challenging, as the outskirts of galaxies are typically faint. Advances in deep imaging surveys improve the detectability of such features, while machine learning enables efficient analysis of large datasets. Using the Pan-STARRS EGIPS catalogue, we develop a deep learning framework by fine-tuning the Zoobot convNext-nano model on 1000 edge-on galaxy FITS images to distinguish warped and non-warped edge-on galaxies with 83\% accuracy. The trained model is then applied to a larger sample, identifying 2088 warped and 1398 non-warped galaxies with a high prediction probability threshold ($\geq 0.85$). Additionally, we use the model to predict on 3226 edge-on galaxies from the Euclid Q1 survey, demonstrating the model's ability to generalise across datasets with differing resolutions. To analyse the model predictions, we employ LayerCAM to identify the regions of galaxy images that contribute to the classification. We find that warped galaxies differ primarily in their structural properties, exhibiting lower axis ratios and higher asymmetry. Warped galaxies were found to be bluer, with younger stellar populations and enhanced star formation. These results highlight the effectiveness of deep learning methods in identifying subtle morphological features, such as warps, and demonstrate their potential for studying structural properties of galaxies in current and upcoming large imaging surveys.

The Fornax galaxy cluster constitutes a suitable laboratory to explore the evolution of galaxies in a dense environment. Recently, the Southern Photometric Local Universe Survey (S-PLUS) has obtained unprecedented photometric information of Fornax, revealing new features regarding its galaxy populations and surrounding regions. In this context, simulations are invaluable tools to interpret the past, present and fate of such observational findings. We aim to deliver a robust photometric catalog of simulated Fornax-like systems in cosmological context, to consistently contrast them with S-PLUS data. We analyze Fornax analogs from the EAGLE and IllustrisTNG simulations, selected using observed properties of the Fornax cluster and its central galaxy NGC 1399. For each system, we generated synthetic photometry in the 12 S-PLUS bands using the SKIRT radiative transfer code, reproducing the instrumental configuration of the S-PLUS survey. Simulated data cubes, mock images, spectral energy distributions, magnitudes and colors were obtained for each galaxy in our selected simulated Fornax analogs. The synthetic photometry and spectra derived from simulations show a good agreement with the S-PLUS observations. We identify particular systems which show some similarity with the spatial distribution of galaxies in Fornax. Such simulated candidates reproduce the observed color-magnitude relation and the spatial substructure between the cluster core and the Fornax A region. Also, simulated galaxies are bluer at higher cluster-centric distances, in agreement with observations. Although modest discrepancies were obtained between the observed and simulated color-magnitude diagrams in some cases, our results support the suitability of our selection criteria and synthetic photometry, and the reliability of current cosmological simulations to reproduce key general features of the Fornax cluster.

Core-collapse supernovae are explosions of massive stars. While most massive stars end as iron-core-collapse supernovae, less massive stars are expected to explode as electron-capture supernovae (ECSNe), defining the low-mass boundary of core-collapse supernovae. ECSNe were proposed $\sim 40$ years ago, and first-principles simulations predict their successful explosions with low energies of $\sim 10^{50}$~erg. Nevertheless, only one convincing candidate, SN~2018zd, has been proposed other than SN~1054, the progenitor of the Crab Nebula. We search for ECSN candidates among Type~II SNe from the literature and a public Zwicky Transient Facility sample, using a color-based diagnostic, selecting ten candidates with blue colors at the middle of the plateau. We classify three as \textit{gold}, for which a spectrum around the middle of the plateau disfavors strong circumstellar-medium interaction that would make the SN bluer, and seven as \textit{silver} without such spectra. Comparing the observed multicolor light curves with radiation-hydrodynamical models, we infer the explosion energies, $(0.4-1.7)\times10^{50}$~erg for the \textit{gold candidates} and $(0.4-2.7)\times10^{50}$~erg including the \textit{silver candidates}, consistent with first-principles predictions and the mass-loss rates, $3\times10^{-3} - 3 \times 10^{-2}~M_{\odot}~{\rm yr}^{-1}$ for the \textit{gold candidates}, which remain similar when the \textit{silver candidates} are included, higher than those expected for the early super-asymptotic-giant-branch phase. The ECSN occurrence ratios among SNe~II are inferred as $3.0^{+10.6}_{-2.9}$ and $15.7^{+17.3}_{-12.7}~\%$ from the \textit{gold} and \textit{silver candidates}, respectively, which we interpret as lower and upper limits. To robustly identify ECSNe and refine this ratio, spectroscopic follow-ups of ECSN candidates around the middle of the plateau are essential.

Wave optics is essential for analyzing lensed gravitational waves (GWs), yet evaluating the diffraction integral $F(ω, y)$ is computationally expensive. We present a Sinusoidal Representation Networks (SIRENs) framework for the dimensionless amplification factor, demonstrating its efficacy and generalization through Point Mass Lens (PML) and Singular Isothermal Sphere (SIS) test cases. Unlike standard architectures that suffer from spectral bias, the network's periodic activation functions structurally align with the integral's oscillatory kernel, effectively resolving high-frequency spectral features. The resulting estimator achieves $\mathcal{O}(10^{-3})$ relative accuracy and a $\sim 100\times$ speedup compared to direct numerical integration. By shifting the computational burden to offline training, our framework yields a stable $\mathcal{O}(1)$ inference complexity. This guarantees constant, sub-millisecond evaluation times even in the weak-lensing diffraction tail where traditional methods stagnate. Additionally, the dimensionless formulation ensures intrinsic scale invariance, enabling direct application across astrophysical regimes from stellar-mass lenses in the ground-based LVK band to supermassive black holes in the space-based LISA band.

Magnetic reconnection in twisted loops has long been invoked as an engine powering energetic transients from black hole accretion to neutron star mergers, yet never directly observed. Here we report the first direct observation of the complete reconnection of this type in a solar flare. We find a magnetic loop twisted to about 540 degrees, far exceeding the 180 degrees twist assumed in existing simulations. This extreme twist inherently enables efficient multiple X-line reconnection, akin to the role of turbulence in contemporary theory. Remarkably, the intertwined end breaks unilaterally after reconnection (unlike symmetric breaking in simulations), forming open field lines that release hot plasma -- providing a promising mechanism for coronal generation or heating. We first detect hard X-ray emission from the current sheet, directly proving it as a particle accelerator. Moreover, we discover a power-law relationship between quasi-periodic oscillation frequency and magnetic field strength across solar flares, black hole binaries, active galactic nuclei, magnetars, and gamma-ray bursts. This relation identifies twisted-pair unilateral reconnection as a common burst mechanism and provides a natural ruler for cosmic magnetic fields. These findings establish an observational foundation for future reconnection theory and simulations, offering a unified framework for magnetically powered bursts.

The BICEP series of experiments has been observing CMB polarisation from the South Pole for over 20 years, with the goal of constraining inflationary gravitational waves. The upcoming data release, using data taken through 2024, is forecasted to constrain the tensor-to-scalar ratio $r$ at the level of $σ(r) \sim 0.005$ (including delensing), with the longer-term goal of reaching $σ(r) \sim 0.001$ by 2034. As the survey sensitivity increases, it is crucial to control instrumental systematics to unprecedented levels, and our goal is to limit dominant sources of systematics to 20% of $σ(r)$ or lower. Achieving this requires careful instrumental characterisation and dedicated end-to-end studies to evaluate the impact of each systematic on cosmological parameters. We first present the BICEP calibration program, which characterises the optical, spectral, and polarisation response of the receivers. We then describe the analysis strategies implemented to identify and mitigate systematic contamination, and we detail simulations used to evaluate the impact of residual effects. We focus in particular on beam systematics, the dominant source of systematics for BICEP receivers. Finally, we report preliminary estimates of the expected level of systematic contamination for upcoming BICEP results, and we discuss approaches to evaluate and mitigate instrumental systematics for future surveys.

The baryonic Tully-Fisher relation (BTFR), a well-established galaxy scaling relation linking the dynamical mass of rotation-supported galaxies through their maximum circular velocity to the baryonic luminous mass, has emerged over the decades as a fundamental scaling relation and as a robust calibrated distance indicator, thereby providing a robust benchmark to test galaxy formation and evolution theories as well as an independent probe of the expansion of the Universe. In this paper, we show for the first time that the BTFR is also simultaneously directly sensitive to the cosmological parameters $Ω_m$ and $σ_8$, the astrophysical feedback from supernovae (SNe) and active galactic nuclei (AGN), and the mass of warm dark matter particles $M_{\rm wdm}$, and can therefore be used as novel, direct probe of cosmology and fundamental physics. We perform simulation-based inference on the large DREAMS cosmological magneto-hydrodynamic simulations suite and train deep neural networks in the form of normalizing flows to estimate the posterior distributions of $Ω_m$, $σ_8$, $M_{\rm wdm}$ and the three astrophysical free parameters, given a BTFR measurement. Our framework is able to recover unbiased values for $Ω_m$ and $σ_8$, with subpercent deviations accuracy and a $\sim 2.6\%$ and $\sim 3.9\%$ median precision, respectively, to capture the warm dark matter particle mass $M_{\rm wdm}$ within a $\sim 30-35\%$ precision, as well as to constrain the SN feedback parameters (but not the one regulating AGN feedback). We conclude that, beyond its usage as a distance indicator and to constrain the baryon cycle and the feedback mechanisms shaping galaxy formation and evolution, the BTFR constitutes a direct independent probe of cosmology and fundamental physics and opens new promising avenues, to be explored with the future Square Kilometer Array.

The study of starburst- and AGN-driven feedback is fundamental for understanding the processes that shape galaxy growth, quench star formation, and drive the co-evolution of galaxies and their central black holes. We present a spatially resolved study of six galaxies at $3 < z < 9$, including starburst- and AGN-dominated systems, observed with JWST/NIRSpec IFU in low- ($R \sim 100$) and high-resolution ($R \sim 2700$) mode. Previous analysis of $R \sim 2700$ data revealed ionized outflows in all these galaxies. We explore possible links between outflows and stellar population properties to assess the mechanisms driving galaxy evolution at early epochs. Combining the spatially resolved stellar population analysis with the outflow properties, we find that the most massive galaxies ($M_\star > 5 \times 10^{10} M_\odot$) with strong AGN-driven ionized outflows show evidence of past quenching episodes occurring within the last $\sim 100$--$300$ Myr, mainly in the nuclear regions ($r < 3$ kpc). In contrast, higher-redshift ($z > 5$) and less massive ($M_\star < 10^{10} M_\odot$) starburst galaxies with powerful (starburst-driven) outflows appear to have experienced continuous growth, with no clear sign of quenching. Massive galaxies with weaker outflows do not show evidence of quenching in their history. One massive AGN host (GS20936) shows evidence for a recent ($10$--$30$ Myr) rejuvenation phase, likely fueled by a recent major merger. These results suggest that (AGN-driven) outflows can already play a key role in shaping the SFHs of massive galaxies at early cosmic epochs.

Relativistic Doppler boosting significantly affects the observed emission of astrophysical jets resulting in observational biases. In this work we investigate the observational biases and modelling opportunities which arise due to relativistic boosting using two X-ray binary case studies. Using the one-sided jet ejecta from MAXI J1535-571, we demonstrate that incorporating non-detections of the receding jet ejecta into kinematic modelling can significantly improve parameter estimation, reducing posterior uncertainties by over 40%. For the bipolar jets of MAXI J1820+070, we recover the intrinsic jet rest-frame emission of both approaching and receding jet components, demonstrating that they follow a common powerlaw evolution. Using this rest-frame emission profile as a base model, we show that current observational strategies strongly bias against detecting ejecta with high initial Lorentz factors >5 and receding ejecta components across a broad region of parameter space. These results highlight the importance of observational strategy selection, particularly early-time and late-time observations, and leveraging non-detections in the modelling of relativistic jets. More generally, quantifying observational biases and maximising modelling capabilities by incorporating the non-detection of receding jets can be employed to enhance interpretation of future gravitational-wave/optically-triggered observations of off-axis, extragalactic jetted transients.

We present a 13-year Chandra/HRC-I study of the M87 jet, using point-spread-function deconvolution to achieve unprecedented sub-arcsecond X-ray resolution of its kiloparsec-scale structure and temporal evolution. The deconvolved images reveal previously blended structures within the jet, including two components in HST-1 and complex morphology in downstream knots. Flux measurements reveal a global decrease in X-ray emission across the jet of up to 84 %, consistent with synchrotron cooling. Modeling the fading yields minimum magnetic field strengths of ~324-1006 $μ$G for HST-1 and ~41-115 $μ$G for knot A. Consistent with synchrotron cooling, multi-wavelength comparisons with ALMA, JWST, and HST show that the principal X-ray structures closely match the jet width and knot locations observed at lower energies, while the X-ray emission is generally shifted upstream. Proper motion measurements show that jet features exhibit both quasi-stationary and superluminal apparent motions, reaching up to 4.8 $c$ for HST-1, and also demonstrate that unresolved component blending can substantially bias inferred velocities. These results demonstrate the unique capability of high-resolution Chandra X-ray imaging over long temporal baselines to resolve the evolving substructure of relativistic jets and to probe the particle acceleration and energy dissipation processes that shape their dynamics.

In an era of highly complex machine learning methods that often are informative but not straightforward to interpret, Principal Component Analysis (PCA) offers a simple, easily interpretable approach. With no fitting parameters, it extracts the most salient statistical trends in data without the need for training sets. In this paper, we explore a large range of composite stellar population models defined for detailed analyses of galaxy spectra from surveys. Six of the most prominent spectral indices are targeted to visualize a PCA-based latent space created by the model data. The age-metallicity degeneracy is broken in the 3-dimensional space spanned by the first three eigenvectors, but we emphasize that non-trivial combinations of all six absorption indices are needed for this. Moreover, the last eigenvector suggests an intriguing tug of war between two Balmer indices: H$γ_A$ and $Hδ_A$, that can help discern the presence of recent bursting behaviour, as it exploits the different behaviour of the two indices over timescales $\sim$0.5-1 Gyr. Comparisons can be made between SDSS and LEGA-C galaxy spectra based on the latent space created by the models. This method, based on pure data, produces excellent results in agreement with standard SPS model fitting techniques, allowing for the study of stellar populations in a variety of surveys or observational/synthetic databases on solid ground.

The orbital histories of the dwarf satellites of the Milky Way (MW) are key to understanding their evolution and placing their present-day properties in a dynamical context. We present the results of the orbit integration of 72 dwarfs in the vicinity of the MW, based on accurate 6D phase-space coordinates from the literature and a suite of six realistic, time-evolving gravitational potentials that account for the mutual interaction of the MW and the Large Magellanic Cloud (LMC). We provide the largest catalogue of orbital parameters for MW dwarfs to date, in terms of both galaxy sample size and range of potentials explored. We also assess the binding status of the dwarfs and estimate their infall times, finding that the majority of them have spent the last 5 Gyr within the MW virial radius. From the reconstructed orbits, we identify ten likely LMC satellites, several of which have experienced very close passages within the LMC stellar disc. For the Small Magellanic Cloud (SMC), we find that its most recent pericentre about the LMC ($\sim$8 kpc, $\sim$170 Myr ago) is consistent with predictions from the direct collision scenario proposed to explain the LMC's offset and tilted bar. We also note a broad temporal coincidence between previous SMC pericentres and star formation rate peaks reported in both Magellanic Clouds, suggesting a causal connection. Finally, we identify Grus II and Tucana IV as possible MW satellites recently captured by the LMC, based on their pronounced orbital deflections and velocities relative to the LMC.

Most observations and clustering analyses suggest that quasars inhabit a narrow range of dark-matter halo masses ($10^{12-13}$ M$_{\odot}$) across cosmic time ($z\lesssim7$). Recent hydrodynamical simulations in gigaparsec-scale volumes now enable direct comparison of this picture with self-consistent galaxy-formation models. We quantify the relation between AGN bolometric luminosity and host halo mass before Cosmic Noon in TNG300, TNG-Cluster, FLAMINGO L1_m8 and L2p8_m9, and in smaller-volume simulations (Illustris, EAGLE, TNG100, and Simba). For AGNs with $L^{\mathrm{AGN}}_{\mathrm{bol}} \ge 10^{42}$ erg s$^{-1}$, more massive haloes host more luminous AGNs on average, but only up to a certain mass. The median luminosity-halo mass relation is highly non-linear, with large scatter, and flattens (FLAMINGO) or turns over (TNG300+TNG-Cluster) at halo mass, $M_{\mathrm{200,crit}} \gtrsim10^{12}$ M$_{\odot}$, at least at $z<5$-6. This high mass AGN quenching also manifests as a characteristic quasar host halo mass: in TNG300+TNG-Cluster, quasars ($L^{\mathrm{AGN}}_{\mathrm{bol}} \sim10^{45-47}$ erg s$^{-1}$) typically reside in haloes of mass $10^{12-12.5}$ M$_{\odot}$ at $z=3$-6, while FLAMINGO quasars extend to median masses of $\sim10^{12.8}$ M$_{\odot}$ at $z\sim3$-4. All simulations predict substantially larger scatter in AGN luminosity at fixed halo mass than in halo mass at fixed luminosity (up to 3 dex versus $\lesssim1$ dex between the 5th and 95th percentiles), implying weak coupling between halo growth and instantaneous SMBH accretion. Consequently, simulated quasar host masses broadly agree with observational estimates. The most luminous AGNs occupy increasingly rare haloes at earlier epochs but typically do not reside in the most massive haloes at any redshift up to $z\approx7$.

We present James Webb Space Telescope (JWST) results of a morphological study of galaxies in the CAnadian NIRISS Unbiased Cluster (CANUCS) and Technicolor surveys, observed in 19 medium- and broadband NIRCam filters in five CANUCS NIRCam Flanking Fields with rest-frame wavelength coverage between $\sim 0.2 - 3.2μm$. Using GALFIT, we measure the morphological parameters of $\sim$ 4,100 star-forming galaxies at $0.6 < z \leq 4$ with stellar masses of $8.5 < \text{log}(M_*/M_\odot) \leq 11.5$. This enables us to concurrently examine how galaxy size varies as a function of stellar mass, redshift, and rest-frame wavelength to provide a novel parametrization of the galaxy size-wavelength relation. Additionally, we analyze the evolution of the galaxy size-mass relation in the rest-frame optical and NIR with the introduction of wavelength as a free parameter. We report a gradient in the slope of the size-mass relation with respect to rest-frame wavelength with a critical crossover mass at $\sim 10^{9.5} M_\odot$. We propose this characteristic mass as the stellar mass at which galaxies transition between diffuse and compact morphologies. We concurrently present the data release of morphological measurements of the five CANUCS-Technicolor NIRCam Flanking Fields in which we provide structural parameters for $\sim$ 41,000 galaxies in up to 29 JWST+HST filters.

The circumgalactic medium (CGM) is the extended multi-phase gaseous halo surrounding a galaxy. To study how its dynamics are shaped by the interplay of radiative cooling and turbulence, we first turn to the TNG50 cosmological magnetohydrodynamical simulation. We analyze the turbulent properties of the CGM of Milky Way-like galaxies at z=0, using a multi-scale filtering technique. We find that the CGM hosts predominantly subsonic flows with Mach number ~ 0.1 - 0.7 that are roughly balanced between compressive and solenoidal components with b = M_comp / M_tot ~ 0.3 - 0.6. We then use these physically motivated properties of turbulent CGM media to set the initial conditions and driving parameters for idealized, turbulent box simulations. These represent ~ 1 kpc^3 volumes of the CGM, but reach resolutions higher than available in cosmological volumes such as TNG50. To do so, we implement a turbulent driving algorithm into the AREPO code based on random but smoothly time-fluctuating acceleration fields, and then run a suite of turbulent boxes across the parameter space, including a model for metal-line radiative cooling. Using these simulations, we study how turbulence seeds density fluctuations that trigger thermal instability and promotes mixing between hot and cold phases, depending on the relative timescales of cooling and mixing. The abundance and evolution of cold ~10^4 K gas depends strongly on the strength of turbulence and the efficiency of cooling.

Gravitational-wave events from mergers of compact objects are a predicted source of high-energy neutrinos. Using data from the IceCube Neutrino Observatory, we search for neutrinos coincident with 85 significant and 945 low-significance gravitational-wave candidate events from compact binary coalescences published in real-time by the LIGO-Virgo-KAGRA collaboration during the first part of its fourth observing run (O4a) and its preceding engineering run, within a time window of $\pm500$ seconds centered on the merger time. We report improvements to the online pipelines, including automatic sending of notices, which has decreased the IceCube real-time response time to gravitational-wave events. In addition, we search for long-duration neutrino emission (up to two weeks after the merger) from three candidate events: two neutron star-black hole mergers, and one low-significance gravitational-wave event with a possible subthreshold gamma-ray counterpart. We use two methods, both of which have been previously used to search for neutrino emission associated with gravitational-wave transients: an unbinned maximum likelihood analysis on significant alerts and a Bayesian analysis accounting for astrophysical priors on both significant and low-significance alerts. We find no statistically significant emission from any of the individual gravitational-wave events analyzed, and set upper limits on the time-integrated flux and energy emitted in high energy neutrinos assuming isotropic emission from each event.