探索全球前沿学术脉络

01.

arXiv (CS.CL) 2026-06-12 DOI: arXiv:2606.12476

Quickest Detection of Hallucination Onset: Delay Bounds and Learned CUSUM Statistics

作者:

Igor Itkin↗

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; at a matched false-alarm rate it detects in 11-13 tokens, against 31 for a linear per-token baseline, and a controlled decomposition attributes most of this advantage 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

阅读与讨论 → 访问原文 →

02.

arXiv (CS.CL) 2026-06-15 DOI: arXiv:2606.14674

AgentSpec: Understanding Embodied Agent Scaffolds Through Controlled Composition

作者:

Jixuan Chen↗Jianzhi Shen↗Haoqiang Kang↗Zhi Hong↗Qingyi Jiang↗Soham Bose↗Yiming Zhang↗Leon Leng↗Amit Vyas↗Lingjun Mao↗Siru Ouyang↗Kun Zhou↗…

LLM agents are increasingly built not as single model calls, but as scaffolded systems that combine reasoning, memory, reflection, action execution, and learning. While such scaffolds often improve performance, they are often embedded in tightly coupled pipelines, making it difficult to isolate component contributions, compare alternative designs, or understand how module interactions shape agent behavior. We introduce AgentSpec, a modular specification framework that represents embodied agents as typed compositions of reusable policy components with standardized interfaces. AgentSpec standardizes the interfaces among perception, memory, reasoning, reflection, action, and optional learning, enabling components to be swapped and recombined under controlled conditions. We instantiate this framework across DeliveryBench, ALFRED, MiniGrid, and RoboTHOR, and analyze reasoning, memory, reflection, and reinforcement-learning modules across model backbones. Our results show that agent performance is governed by scaffold compatibility and interaction effects rather than isolated module strength. In particular, structured multi-granularity memory improves long-horizon state tracking, reasoning and memory interact non-uniformly across environments, reflection trades off correction and cost, and RL-trained policies compose best when optimized with deployment-time scaffold structure. AgentSpec provides a controlled foundation for studying, comparing, and designing composable LLM agents. Our code, baselines and interactive playground are publicly available at https://agentspec-embodied.github.io.

阅读与讨论 → 访问原文 →

03.

arXiv (CS.AI) 2026-06-11 DOI: arXiv:2606.04145

EvalStop: Using World Feedback to Detect and Correct Reward Overoptimization in Multi-Tenant RLHF Platforms

作者:

Guilin Zhang↗Chuanyi Sun↗Shahryar Sarkani↗John M. Fossaceca↗

arXiv:2606.04145v2 Announce Type: replace-cross Abstract: Cloud LLM fine-tuning platforms increasingly serve RLHF workloads, where a learned reward model is optimized as a proxy for human quality. As Gao et al. (2023) showed, this proxy diverges from world feedback (downstream eval metrics) under sustained optimization pressure, a phenomenon known as reward overoptimization. Existing platform schedulers ignore this divergence: non-clairvoyant schedulers optimize JCT without any quality signal, SLAQ-style quality-aware schedulers use training loss (a weaker proxy that drops monotonically through hacking), and classical per-job early stopping requires human monitoring and does not free shared GPUs. We propose EvalStop, a composable scheduling primitive that terminates jobs on k consecutive eval-score declines, releases GPUs, preserves the best checkpoint, and delegates to any base scheduler. We frame scheduler-level early stopping as a detection problem and evaluate it in a discrete-event simulator whose RLHF workload mixes reward-hacking and structurally healthy runs, with ground-truth labels hidden from schedulers. On RLHF-heavy workloads (80% RLHF, 64 GPUs), EvalStop achieves precision 98% / recall 99% / FPR 1.5% while improving JCT by 9% and cutting wasted compute by 22% over SRTF-Est (p

阅读与讨论 → 访问原文 →

04.

Nature (Science) 2026-06-17 DOI: HASH:8f0fbd8d715b37a6817e31f5ab9e9f4e

Structure of the pre-initiation complex explains CMGE biogenesis

作者:

Thomas Pühringer↗

When cells enter S phase, bidirectional DNA replication is initiated through the kinase-regulated recruitment of three activators (Cdc45, GINS and Pol ε) to a duplex-DNA-loaded double hexamer of minichromosome maintenance (MCM) ATPases. Together, these proteins form two CMGE helicases that establish divergent replication forks as they become separated1. Here, to gain an understanding of CMGE biogenesis, we reconstituted the pre-initiation complex with purified yeast proteins. The cryo-electron-microscopy structure shows a set of firing factors caught in the act of assembling two symmetrical CMGEs. We show how stepwise complex formation reshapes MCM in preparation for DNA opening, and we explain how ATP promotes firing-factor ejection and CMGE maturation. We find that although Sld2 facilitates the recruitment of GINS to MCM, as expected, it also aids the efficient separation of the CMGE dimer, and is essential for the ejection of the lagging strand from MCM. These findings have direct implications for our understanding of the metazoan Sld2 orthologue, RECQL4, and point to a replication-fork establishment mechanism that is conserved across eukaryotes. Cryo-electron microscopy and biochemical reconstitution experiments in yeast provide insight into the assembly of the CMGE complex, a helicase that establishes bidirectional DNA replication in eukaryotic cells, and elucidate the role of the firing factor Sld2.

阅读与讨论 → 访问原文 →

05.

arXiv (CS.CV) 2026-06-11 DOI: arXiv:2511.16672

EvoLMM: Self-Evolving Large Multimodal Models with Continuous Rewards

作者:

Omkar Thawakar↗Shravan Venkatraman↗Ritesh Thawkar↗Abdelrahman Shaker↗Hisham Cholakkal↗Rao Muhammad Anwer↗Salman Khan↗Fahad Khan↗

Recent advances in large multimodal models (LMMs) have enabled impressive reasoning and perception abilities, yet most existing training pipelines still depend on human-curated data or externally verified reward models, limiting their autonomy and scalability. In this work, we strive to improve LMM reasoning capabilities in a purely unsupervised fashion (without any annotated data or reward distillation). To this end, we propose a self-evolving framework, named EvoLMM, that instantiates two cooperative agents from a single backbone model: a Proposer, which generates diverse, image-grounded questions, and a Solver, which solves them through internal consistency, where learning proceeds through a continuous self-rewarding process. This dynamic feedback encourages both the generation of informative queries and the refinement of structured reasoning without relying on ground-truth or human judgments. When using the popular Qwen2.5-VL as the base model, our EvoLMM yields consistent gains upto $\sim$3\% on multimodal math-reasoning benchmarks, including ChartQA, MathVista, and MathVision, using only raw training images. We hope our simple yet effective approach will serve as a solid baseline easing future research in self-improving LMMs in a fully-unsupervised fashion. Our code and models are available at https://github.com/mbzuai-oryx/EvoLMM.

阅读与讨论 → 访问原文 →

06.

arXiv (CS.AI) 2026-06-12 DOI: arXiv:2601.06572

Hellinger Multimodal Variational Autoencoders

作者:

Huyen Vo↗Isabel Valera↗

arXiv:2601.06572v4 Announce Type: replace-cross Abstract: Multimodal variational autoencoders (VAEs) are widely used for weakly supervised generative learning with multiple modalities. Predominant methods aggregate unimodal inference distributions using either a product of experts (PoE), a mixture of experts (MoE), or their combinations to approximate the joint posterior. In this work, we revisit multimodal inference through the lens of probabilistic opinion pooling, an optimization-based approach. We start from Hölder pooling with $\alpha=0.5$, which corresponds to the unique symmetric member of the $\alpha-divergence$ family, and derive a moment-matching approximation, termed Hellinger. We then leverage such an approximation to propose HELVAE, a multimodal VAE that avoids sub-sampling, yielding an efficient yet effective model that: (i) learns more expressive latent representations as additional modalities are observed; and (ii) empirically achieves better trade-offs between generative coherence and quality, outperforming state-of-the-art multimodal VAE models.

阅读与讨论 → 访问原文 →

07.

arXiv (CS.AI) 2026-06-18 DOI: arXiv:2602.06774

Towards Understanding What State Space Models Learn About Code

作者:

Jiali Wu↗Abhinav Anand↗Shweta Verma↗Mira Mezini↗

arXiv:2602.06774v2 Announce Type: replace Abstract: State Space Models (SSMs) have emerged as an efficient alternative to the Transformer architecture. Prior work shows that, when trained under comparable conditions, SSMs can match or surpass Transformers on code understanding tasks. However, their internal mechanisms remain a black box. We present the first systematic analysis of what SSM-based code models learn along with the direct comparison between SSM and Transformer models in this domain. Our analysis shows that SSMs capture syntactic and semantic structure more effectively than Transformers during pretraining but forgets certain relations during fine-tuning on some tasks. To investigate this behavior, we introduce SSM-Interpret, a frequency-domain framework that exposes a spectral shift toward short-range dependencies during fine-tuning. Guided by these findings, we propose architectural modifications that significantly improve the performance of SSM-based code model by upto +6 MRR on NLCodeSearch. This demonstrates that our analysis not only explains model behavior but also leads directly to better designs.

阅读与讨论 → 访问原文 →

08.

arXiv (CS.CV) 2026-06-15 DOI: arXiv:2606.14686

CottonLeafVision: An Explainable and Robust Deep Learning Framework for Cotton Leaf Disease Classification

作者:

Rafi Ahamed↗Md. Abir Rahman↗Tasnia Tarannum Roza↗Munaia Jannat Easha↗Md. Asif Khan↗Sudeepta Mandal↗

Globally, cotton is a highly economically beneficial crop, as the textile industry heavily depends on it. So, the precise identification and detection of cotton leaf disease is crucial for economic stability. The development goal of "CottonLeafVision" is to accurately classify and detect cotton leaf disease. With this goal, we have evaluated multiple pretrained Deep Convolutional Neural Networks, including DenseNet201, InceptionV3, and VGG19 on a publicly available cotton leaf disease image dataset. This image dataset includes seven classes, six disease classes, and one healthy class, collected under various field conditions reflecting real-world challenges. Among these pretrained models, with DenseNet201, we have achieved the highest classification accuracy of 98%. To enhance the model reliability and interpretability, we have implemented different techniques and methods such as Gradient-weighted Class Activation Mapping (Grad-CAM), occlusion sensitivity analysis and adversarial training to increase the noise resistance of the model. Finally, we have developed a prototype in order to utilize the model's capabilities on real life agriculture. This paper shows the deep learning model's capabilities to classify the disease in real-life cotton disease management situations.

阅读与讨论 → 访问原文 →

09.

arXiv (CS.CV) 2026-06-17 DOI: arXiv:2606.17463

WeaveLA: Event Driven Cross-Subtask Latent Memory Weaving for Repetitive Robot Manipulation

作者:

Shoujing Zhu↗Zhenyang Liu↗Fungmiu Wang↗Jiafeng Wang↗Bo Yue↗Guiliang Liu↗Simo Wu↗Xiangyang Xue↗Taiping Zeng↗

Vision-Language-Action (VLA) policies have achieved remarkable single-step manipulation, yet they remain brittle precisely where each stage depends on what was just completed. The core issue is structural: short-window VLAs lack an explicit channel for rouxting information across sub-task boundaries, and existing memory-augmented variants either write at every frame, retrieve from demonstration-time stages, or fire at sub-goal events without performing an explicit sub-task-to-sub-task hand-off into the action expert. We identify the sub-goal completion event as the natural temporal unit for cross-subtask memory hand-off, and present WeaveLA (Weave Latent memory for Vision-Language-Action policies), a cross-subtask memory interface that, on top of a frozen VLA backbone, compresses each completed segment into latent tokens via query-driven attention pooling and routes them directly into the action-generation path of the next sub-task. This event-triggered, action-side design preserves the base policy's short-window interface while adding a lightweight cross-subtask channel. Through stratified evaluation on RoboMME with a $\pi_{0.5}$ backbone, WeaveLA's gains land exactly where the channel is needed: on the hardest repetition slice (SwingXtimes, $N{=}3$), success rises from $0\%$ to $47.8\%$, while single-execution episodes remain unchanged. Per-episode paired analysis confirms the gains are confined to tasks whose causal structure requires cross-subtask information.

阅读与讨论 → 访问原文 →

10.

arXiv (math.PR) 2026-06-12 DOI: arXiv:2606.13575

Dimension-free Markov–Bernstein inequalities for product measures

作者:

Egor Kosov↗

arXiv:2606.13575v1 Announce Type: cross Abstract: We study dimension-free Markov–Bernstein inequalities for polynomials with respect to product probability measures. In the Gaussian case, for $p\ge4$, we prove that \[ \|\nabla f\|_{L^p(\gamma^n)} \le C(p)d^{\frac12+\theta_p} \|f\|_{L^p(\gamma^n)} \] for every polynomial $f$ of degree at most $d$, where $\theta_p\le \frac{2}{3p}$ and $\theta_p=0$ whenever $p$ is an even integer. Thus, for even integer exponents, we establish the sharp dependence on the degree conjectured by Eskenazis–Ivanisvili. For general $p\ge4$, the estimate improves upon their dimension-free inequality. We also obtain dimension-free Markov–Bernstein inequalities with sharp dependence on the degree for even integer exponents beyond the Gaussian setting. We first prove such estimates for the uniform distribution on the unit cube and then extend them to products of absolutely continuous measures with unimodal densities. Finally, we treat products of one-dimensional Freud measures with densities proportional to $e^{-|t|^{2m}}$.

阅读与讨论 → 访问原文 →

11.

arXiv (CS.AI) 2026-06-16 DOI: arXiv:2606.16541

The Faithfulness Gap: Certifying Semantic Equivalence Between Natural-Language and Formal Mathematical Statements

作者:

Noor Islam S. Mohammad↗Tamim Sheikh↗

arXiv:2606.16541v1 Announce Type: new Abstract: Autoformalization, translating natural-language mathematics into formal proof assistants, is bottlenecked not by translation fluency but by faithfulness: a formal statement can typecheck and be provable, yet still encode a different theorem than the source intended. We introduce Bidirectional Provability Fingerprinting (\bpf{}), a framework that certifies faithfulness by characterizing each candidate through its forward and backward consequence neighborhoods in the ambient theory and matching these against probes derived from the natural-language statement. We further introduce four novel components: (i) Counterfactual Probe Generation (\cpg{}), a contrastive procedure that synthesizes probes targeting specific drift directions; (ii) the Equivalence Spectrum, a continuous faithfulness score that replaces brittle binary verdicts; (iii) Adaptive Probe Budget Allocation (\apba{}), an information-theoretic budget router; and (iv) Faithfulness-Guided Decoding (\fgd{}), which uses \bpf{} signals as a reward during autoformalization. We prove a drift detection theorem and a PAC-faithfulness result establishing that the equivalence class of a natural language statement is learnable from $\mathcal{O}(\log(1/\delta)/\varepsilon)$ probes under mild assumptions. We release \driftbench{}, a benchmark of $2{,}183$ NL/Lean~4 pairs with controlled drift labels across six subfields of mathlib4. \bpf{}\,+\,\cpg{} detects $89.6\%$ of drifted formalizations at a $3.0\%$ false-positive rate-against $41.2\%$ for typecheck and $63.3\%$ for LLM-judge baselines, and \fgd{} reduces the rate at which a state-of-the-art autoformalizer emits drifted statements by $47\%$. https://pmlrbd.github.io/BPF/

阅读与讨论 → 访问原文 →

12.

arXiv (CS.LG) 2026-06-15 DOI: arXiv:2604.04611

Dynamic Free-Rider Detection in Federated Learning via Simulated Attack Patterns

作者:

Motoki Nakamura↗

arXiv:2604.04611v2 Announce Type: replace Abstract: Federated learning (FL) enables multiple clients to collaboratively train a global model by aggregating local updates without sharing private data. However, FL often faces the challenge of free-riders, clients who submit fake model parameters without performing actual training to obtain the global model without contributing. Chen et al. proposed a free-rider detection method based on the weight evolving frequency (WEF) of model parameters. This detection approach is a leading candidate for practical free-rider detection methods, as it requires neither a proxy dataset nor pre-training. Nevertheless, it struggles to detect ``dynamic'' free-riders who behave honestly in early rounds and later switch to free-riding, particularly under global-model-mimicking attacks such as the delta weight attack and our newly proposed adaptive WEF-camouflage attack. In this paper, we propose a novel detection method S2-WEF that simulates the WEF patterns of potential global-model-based attacks on the server side using previously broadcasted global models, and identifies clients whose submitted WEF patterns resemble the simulated ones. To handle a variety of free-rider attack strategies, S2-WEF further combines this simulation-based similarity score with a deviation score computed from mutual comparisons among submitted WEFs, and separates benign and free-rider clients by two-dimensional clustering and per-score classification. This method enables dynamic detection of clients that transition into free-riders during training without proxy datasets or pre-training. We conduct extensive experiments across three datasets and five attack types, demonstrating that S2-WEF achieves higher robustness than existing approaches.

阅读与讨论 → 访问原文 →

13.

arXiv (CS.CL) 2026-06-12 DOI: arXiv:2602.09379

LingxiDiagBench: A Multi-Agent Framework for Benchmarking LLMs in Chinese Psychiatric Consultation and Diagnosis

作者:

Shihao Xu↗Tiancheng Zhou↗Jiatong Ma↗Yanli Ding↗Yiming Yan↗Ming Xiao↗Guoyi Li↗Haiyang Geng↗Yunyun Han↗Jianhua Chen↗Yafeng Deng↗

Mental disorders are highly prevalent worldwide, but the shortage of psychiatrists and the inherent subjectivity of interview-based diagnosis create substantial barriers to timely and consistent mental-health assessment. Progress in AI-assisted psychiatric diagnosis is constrained by the absence of benchmarks that simultaneously provide realistic patient simulation, clinician-verified diagnostic labels, and support for dynamic multi-turn consultation. We present LingxiDiagBench, a large-scale multi-agent benchmark that evaluates LLMs on both static diagnostic inference and dynamic multi-turn psychiatric consultation in Chinese. At its core is LingxiDiag-16K, a dataset of 16,000 EMR-aligned synthetic consultation dialogues designed to reproduce real clinical demographic and diagnostic distributions across 12 ICD-10 psychiatric categories. Through extensive experiments across state-of-the-art LLMs, we establish key findings: (1) although LLMs achieve high accuracy on binary depression–anxiety classification (up to 92.3%), performance deteriorates substantially for depression–anxiety comorbidity recognition (43.0%) and 12-way differential diagnosis (28.5%); (2) dynamic consultation often underperforms static evaluation, indicating that ineffective information-gathering strategies significantly impair downstream diagnostic reasoning; (3) consultation quality assessed by LLM-as-a-Judge shows only moderate correlation with diagnostic accuracy, suggesting that well-structured questioning alone does not ensure correct diagnostic decisions. We release LingxiDiag-16K and the full evaluation framework to support reproducible research at https://github.com/Lingxi-mental-health/LingxiDiagBench.

阅读与讨论 → 访问原文 →

14.

arXiv (CS.AI) 2026-06-19 DOI: arXiv:2604.13416

DF3DV-1K: A Large-Scale Dataset and Benchmark for Distractor-Free Novel View Synthesis

作者:

Cheng-You Lu↗Yi-Shan Hung↗Wei-Ling Chi↗Hao-Ping Wang↗Charlie Li-Ting Tsai↗Yu-Cheng Chang↗Yu-Lun Liu↗Thomas Do↗Chin-Teng Lin↗

arXiv:2604.13416v2 Announce Type: replace-cross Abstract: Advances in radiance fields have enabled photorealistic novel view synthesis. In several domains, large-scale real-world datasets have been developed to support comprehensive benchmarking and to facilitate progress beyond scene-specific reconstruction. However, for distractor-free radiance fields, a large-scale dataset with clean and cluttered images per scene remains lacking, limiting the development. To address this gap, we introduce DF3DV-1K, a large-scale real-world dataset comprising 1,048 scenes, each providing clean and cluttered image sets for benchmarking. In total, the dataset contains 89,924 images captured using consumer cameras to mimic casual capture, spanning 128 distractor types and 161 scene themes across indoor and outdoor environments. A curated subset of 41 scenes, DF3DV-41, is systematically designed to evaluate the robustness of distractor-free radiance field methods under challenging scenarios. Using DF3DV-1K, we benchmark nine recent distractor-free radiance field methods and 3D Gaussian Splatting, identifying the most robust methods and the most challenging scenarios. Beyond benchmarking, we demonstrate an application of DF3DV-1K by fine-tuning a diffusion-based 2D enhancer to improve radiance field methods, achieving average improvements of 0.96 dB PSNR and 0.057 LPIPS on the held-out set (e.g., DF3DV-41) and the On-the-go dataset. We hope DF3DV-1K facilitates the development of distractor-free vision and promotes progress beyond scene-specific approaches. The dataset and leaderboard are available at https://johnnylu305.github.io/df3dv1k_web/.

阅读与讨论 → 访问原文 →

15.

medRxiv (Medicine) 2026-06-22 DOI: HASH:6210cc077724c79998a743fa9b0d65f1

Generative Artificial Intelligence in Psychotherapy Practice: A Global Online Survey of Mental Health Professionals' Adoption

作者:

Blease↗Hagström↗Gaab↗Carey↗Cipriani↗Gorman↗Nascimento↗A. F↗Fitzgerald↗Holtz↗Schwarz↗Strudwick↗…

Background: Generative artificial intelligence (GenAI) tools, including large language model (LLM)-based platforms such as ChatGPT, Google Gemini, and Microsoft Copilot, are being adopted across healthcare settings with increasing speed. Despite the increasing popularity of GenAI, empirical data on the extent and nature of adoption by mental health clinicians in routine psychotherapy practice globally remain scarce. Objective: This study aimed to characterize current use patterns of GenAI tools among a global sample of practicing mental health professionals, including prevalence of use, specific tools employed, clinical and administrative purposes served, perceived effect on workload, and the institutional context shaping adoption (e.g., encouragement, prohibition, and training). Methods: We administered a cross-sectional online survey to a global convenience sample of licensed mental health professionals who provide psychotherapy as part of the scope of their practice (i.e., psychotherapists, psychologists, counsellors, nurses, and psychiatrists). Participants were recruited via professional networks, purposely avoiding the use of social media platforms. Within the survey, we captured GenAI use behaviors in psychotherapy contexts, and demographic and professional background data. Descriptive statistics were analyzed for all variables. Multivariate logistic regression was used to examine demographic and professional predictors of GenAI use. Results: A total of 766 mental health professionals who provide psychotherapy from 30 countries completed the survey. Of these, 54.6% (n=418) reported having purposely used at least one GenAI tool in psychotherapy clinical practice. ChatGPT was the most frequently used tool (354/418, 84.7%). The most commonly reported clinical purpose was assisting with treatment planning (175/418, 41.9%), followed by managing administrative tasks (173/418, 41.4%) and generating psychoeducational materials for clients (166/418, 39.7%). 82.8% of AI users reported that these tools reduced their overall work burden. Only 18.1% (139/766) of respondents reported institutional encouragement to use AI tools, while 81.1% (621/766) reported not having received any professional training on AI use. Predictors of AI adoption included younger age and rural practice setting. Conclusions: In this global convenience sample survey, GenAI use among mental health professionals in psychotherapy settings is widespread, concentrated in a wide variety of clinical and administrative tasks. Formal training and institutional guidance substantially lag behind current adoption patterns. These findings highlight an urgent need for evidence-based competency frameworks, regulatory clarity, and professional education to support safe and ethically informed integration of AI into clinical mental health practice.

阅读与讨论 → 访问原文 →

16.

arXiv (CS.CL) 2026-06-19 DOI: arXiv:2606.19625

Where Does Social Reasoning Come From? Capability Provenance in Language Models

作者:

Glenn Matlin↗Chandreyi Chakraborty↗Saehee Eom↗Mika Okamoto↗Rayan Castilla↗Louis Jaburi↗Alvin Deng↗Taywon Min↗Lucia Quirke↗Stella Biderman↗Mark Riedl↗

We use training-data attribution as an interpretable tool for capability discovery, mapping which regions of the pretraining corpus support social-reasoning versus STEM-reasoning in OLMo3-7B. Training-data attribution measures how strongly each training document influences a model's predictions on a benchmark, but document-level scores are too noisy to identify which corpus regions support which capabilities, and prior work has emphasized factual knowledge rather than reasoning. We compute gradient-based attribution (TrackStar via Bergson) over a working set drawn from the de-duplicated Dolma3 mix, aggregate influence across WebOrganizer's 24-format x 24-topic taxonomy (576 bins), and contrast benchmark pairs in a 2x2 design that varies domain (social vs. STEM) and capability type (reasoning vs. knowledge): SocialIQA and MMLU Social Sciences against ARC-Challenge and MMLU STEM. Social and STEM reasoning draw on qualitatively distinct corpus regions, and the contrast is sharper at the reasoning level than at the knowledge level. Targeted machine unlearning provides partial causal validation: forgetting high-attribution topic bins (e.g., Literature for SocialIQA) degrades the aligned benchmark more than within-bin random baselines, and we open-source all code, sampling manifests, the bin-level influence matrix, and unlearning checkpoints.

阅读与讨论 → 访问原文 →

17.

medRxiv (Medicine) 2026-06-15 DOI: HASH:09934c3d6b0d32a1f47cd32b933431d5

Excitation-Inhibition Balance in Schizophrenia Spectrum Disorders: EEG Criticality Reflects Frontal Metabolites and a Potential Compensatory Mechanism

作者:

Hasanaj↗Kallweit↗M. S↗Karsli↗Meisinger↗Boudriot↗Roell↗Melcher↗Vural↗Schulz↗Klimas↗Schmoelz↗…

Background The excitation-inhibition (E-I) balance is essential for normal brain functioning, while deviations from this balance have been implicated in several psychiatric disorders. However, the extent to which electroencephalography (EEG) and proton magnetic resonance spectroscopy (1H-MRS) E-I markers are altered in schizophrenia spectrum disorders (SSD), how they converge across modalities, and how they relate to cognitive performance and clinical symptoms remain insufficiently characterized. Methods We recruited 111 healthy controls (HC) and 113 individuals with SSD. All participants underwent resting-state EEG and 1H-MRS. Metabolites were measured either in the anterior cingulate cortex (ACC; NSSD = 63, NHC = 58) or in the left dorsolateral prefrontal cortex (lDLPFC; NSSD = 50, NHC = 53), from which gamma-aminobutyric acid (GABA), glutamate + glutamine (Glx), and the Glx/GABA ratio were extracted. Extracted EEG E-I markers included oscillatory activity, aperiodic activity, functional E-I, microstates, multiscale entropy, and neuronal avalanche criticality. Results MRS results showed no group differences in GABA, Glx, or the Glx/GABA ratio. In contrast, most EEG-derived E-I markers indicated increased cortical inhibition in SSD, including steeper aperiodic exponents, prolonged microstate durations, and greater prevalence of subcritical states. However, functional E-I showed a divergent pattern, suggesting balanced dynamics in SSD and relatively inhibition-weighted dynamics in HC. Across groups, higher ACC and lDLPFC GABA predicted a lower kappa index, whereas a higher lDLPFC Glx/GABA ratio was associated with a higher kappa index. In SSD, reduced avalanche criticality was associated with better cognition and less severe symptoms. Conclusion Several EEG-derived E-I proxies, but not MRS measures, indicate an increased cortical inhibition in SSD. Criticality indices best capture frontal neurochemical metabolites and improvements in clinical symptoms, potentially reflecting inhibitory compensation mechanisms in SSD.

阅读与讨论 → 访问原文 →

18.

arXiv (CS.AI) 2026-06-16 DOI: arXiv:2606.14710

Poster: EdgeCitadel – Hybrid NATS-MQTT Orchestration for Edge Multi-Agent Systems

作者:

Zhonghao Zhan↗Yefan Zhang↗Hamed Haddadi↗

arXiv:2606.14710v1 Announce Type: cross Abstract: Edge-resident AI agents increasingly span home servers, IoT hubs, laptops, and phones, yet their coordination stacks still assume cloud-style transports or a central relay. We present EdgeCitadel, an edge multi-agent orchestration platform built around a single NATS 2.10 server with the built-in MQTT adapter. The design combines MQTT connectivity for heterogeneous agents, JetStream-backed persistence and replay for backend services, direct peer delegation over a shared subject namespace, and a passive aggregator that visualizes and stores traffic without sitting on the delivery path. Our poster highlights the migration from MQTT relay prototypes (common in IoT communication) to the current hybrid architecture and demonstrates a working cross-device testbed spanning ARM64, x64, and Android clients.

阅读与讨论 → 访问原文 →

19.

medRxiv (Medicine) 2026-06-16 DOI: HASH:09c2f5e2407b10ffb1f04328fce24c48

Validating an Early Pregnancy HbA1c as the Screening Test for Gestational Diabetes Mellitus: Findings from PRISMA Pakistan Cohort

作者:

Naz↗Yazdani↗Hoodbhoy↗Ghebremihael-Weldeselassie↗Mazhar↗Hotwani↗Jehan↗Nisar↗M. I↗Iqbal↗

Background: Early identification of gestational diabetes mellitus (GDM) is critical to improving maternal and neonatal outcomes, particularly in resource-constrained settings where universal oral glucose tolerance testing (OGTT) is burdensome. We assessed whether early-pregnancy HbA1c alone or combined with common risk factors can predict GDM and reduce the burden of OGTT requirements in a peri-urban cohort in Karachi, Pakistan. Methods: We conducted a secondary analysis of the Pregnancy Risk Infant Surveillance and Measurement Alliance (PRISMA) Pakistan cohort. Women enrolled before 20 weeks' gestation with available early-pregnancy HbA1c and a 2-hour 75g OGTT at 24 to 28 weeks were included. We externally validated GDM prediction models originally developed in the STRiDE-India cohort. Model performance was evaluated using receiver operating characteristic (ROC) curves and area under the curve (AUC). We assessed four models: HbA1c alone (Model 1a); age, BMI, and family history of diabetes mellitus (FH DM) (Model 1b); HbA1c combined with age, BMI, and FH DM (Model 2); and an extended model, i.e., Model 2 combined with socioeconomic status, gestational age, parity, systolic and diastolic blood pressure (Model 3). A dual-threshold approach was applied to assess rule-in and rule-out performance. Results: Among 2,489 women, GDM incidence was 7.5% (n=186). Models with a broader set of predictors demonstrated higher AUC values, with Model 2 achieving an AUC of 0.61 (95% CI: 0.57, 0.66). Including additional factors (Model 3) did not further improve predictive ability (AUC: 0.62; 95% CI: 0.58, 0.66). In addition, at predefined thresholds, Model 2 achieved sensitivity of 73.7% (rule-out) and specificity of 83.5% (rule-in), with the potential to reduce OGTT requirements (58.5%). Conclusions: Early-pregnancy risk stratification using HbA1c combined with simple clinical predictors offers a pragmatic approach to streamline GDM screening among high-risk pregnant women. A dual-threshold strategy using Model 2 could reduce reliance on universal OGTT while prioritizing high-risk women for confirmatory testing.

阅读与讨论 → 访问原文 →

20.

medRxiv (Medicine) 2026-06-10 DOI: HASH:3c6d6288c2a3e7e876677ea820b7faed

Development of an Open-Access Action Observation Video Library for Upper Limb Motor Rehabilitation

作者:

Madison↗Wheaton↗L. A↗Rowe↗

Background: Occupational therapists can improve stroke survivors hand and arm movement and participation in daily activities through action observation (AO). AO involves watching another persons hand or arm complete a movement or task. While research generally supports the use of AO with stroke survivors, there are limited AO videos are available to occupational therapists which makes applying AO challenging. Objective: The purpose of this work is to develop structured and widely accessible tool to support access to AO for stroke survivors, occupational therapists, and researchers. Methods: To develop an AO video library for stroke rehabilitation, functional and non-functional upper limb task deficits were first identified through clinical observations and clinician interviews to establish a prioritized list of daily activities. In collaboration with media production specialists, healthy adult volunteers were recruited and filmed performing these tasks from both first- and third-person perspectives. The recorded videos were then systematically edited, enhanced with instructional title slides, and distributed via a public YouTube channel for clinical application and a categorized digital repository for research purposes. Results: Initial assessments revealed a complete lack of familiarity, awareness, and utilization of AO resources among local occupational therapists, despite high perceived clinical utility. To address this gap, a final library of 150 tasks was established, resulting in the production of 419 finalized, standardized videos featuring six healthy volunteers. For clinical application, these videos were hosted on a free, public YouTube channel organized into 18 functional playlists, while a parallel set was structured into distinct movement categories for research repository storage. Conclusion: By providing a structured and highly accessible tool, this repository enables clinicians, researchers, and caregivers to readily implement evidence-based action observation interventions in both clinical and home settings.

阅读与讨论 → 访问原文 →

21.

arXiv (CS.CL) 2026-06-16 DOI: arXiv:2606.16817

Understanding the Behaviors of Environment-aware Information Retrieval

作者:

Ruifeng Yuan↗Chaohao Yuan↗David Dai↗Yu Rong↗Hong Cheng↗Hou Pong Chan↗Chenghao Xiao↗

Recent retrieval-augmented generation (RAG) approaches have demonstrated strong capability in handling complex queries, yet current research overlooks a critical challenge: different retrievers require fundamentally different query formulation strategies for optimal performance. In this work, we present the first systematic analysis of how LLMs can learn to adapt their query formulation strategies for different retrievers via reinforcement learning (RL). Our empirical study reveals that RL effectively teaches an LLM to tailor its queries to specific retriever characteristics. We discover that different retrievers exhibit surprisingly distinct optimal query styles (e.g., descriptive vs. question-like), suggesting strategies learned for one retriever ineffective for another. We further show that performance can be enhanced by incorporating retriever-specific human guidance and by scaling model size. To facilitate learning over multi-retrieval-step trajectories, we introduce a branching-based rollout technique that improves training stability. Our work provides the first empirical evidence and actionable insights for building truly retriever-aware RAG systems. Code and resources are available at https://github.com/LCO-Embedding/Envs-aware-Information-Retrieval.

阅读与讨论 → 访问原文 →

22.

arXiv (CS.AI) 2026-06-16 DOI: arXiv:2603.10384

Beyond Scalars: Evaluating and Understanding LLM Reasoning via Geometric Progress and Stability

作者:

Xinyan Jiang↗Ninghao Liu↗Di Wang↗Lijie Hu↗

arXiv:2603.10384v3 Announce Type: replace Abstract: Evaluating LLM reliability via scalar probabilities often fails to capture the structural dynamics of reasoning. We introduce TRACED, a framework that assesses reasoning quality through theoretically grounded geometric kinematics. By decomposing reasoning traces into Progress (displacement) and Stability (curvature), we reveal a distinct topological divergence: correct reasoning manifests as high-progress, stable trajectories, whereas hallucinations are characterized by low-progress, unstable patterns (stalled displacement with high curvature fluctuations). Leveraging these signatures, our probabilistic framework achieves competitive performance and superior robustness across diverse benchmarks. Crucially, TRACED bridges geometry and cognition by mapping high curvature to ''Hesitation Loops'' and displacement to ''Certainty Accumulation'', offering a physical lens to decode the internal dynamics of machine thought.

阅读与讨论 → 访问原文 →

23.

arXiv (CS.LG) 2026-06-17 DOI: arXiv:2410.08562

Adaptable Method for Crystal Design across Diverse Constraints and Objectives with Pretrained Property Predictors

作者:

Akihiro Fujii↗Yoshitaka Ushiku↗Koji Shimizu↗Anh Khoa Augustin Lu↗Satoshi Watanabe↗

arXiv:2410.08562v5 Announce Type: replace-cross Abstract: Advanced crystal design can accelerate materials discovery across applications from photovoltaics to spintronics. Practical design must satisfy multiple properties and physical constraints, yet existing machine-learning-based approaches to such design often depend on large datasets, retraining, or task-specific generators. Here, we show that direct predictor-guided gradient optimization enables data-efficient, constraint-rich crystal design by combining off-the-shelf predictors with site-wise element masks, template initialization, and task-specific losses. In perovskites, it outperformed generative and Bayesian baselines under three targets – band gap, formation energy, and tolerance factor – and two hard constraints. DFT assessment further showed band-gap targeting competitive with a leading generative model despite using predictors trained on roughly one-tenth of the data. By flexibly combining pretrained predictors with application-oriented masks and custom losses, the same framework supported half-metal design. Such modularity could help researchers and engineers translate diverse application requirements directly into optimized candidate crystals with minimal computational cost.

阅读与讨论 → 访问原文 →

24.

arXiv (CS.CV) 2026-06-11 DOI: arXiv:2604.13326

Right Regions, Wrong Labels: Semantic Label Flips in Segmentation under Correlation Shift

作者:

Akshit Achara↗Yovin Yahathugoda↗Nick Byrne↗Michela Antonelli↗Esther Puyol Anton↗Alexander Hammers↗Andrew P. King↗

The robustness of machine learning models can be compromised by spurious correlations between non-causal features in the input data and target labels. A common way to test for such correlations is to train on data where the label is strongly tied to some non-causal cue, then evaluate on examples where that tie no longer holds. This idea is well established for classification tasks, but for semantic segmentation the specific failure modes are not well understood. We show that a model may achieve reasonable overlap while assigning the wrong semantic label, swapping one plausible foreground class for another, even when object boundaries are largely correct. We focus on this semantic label-flip behaviour and quantify it with a simple diagnostic (Flip) that counts how often ground truth foreground pixels are assigned the wrong foreground identity while remaining predicted as foreground. In a setting where category and scene are correlated during training, increasing the correlation consistently widens the gap between common and rare test conditions and increases these within-object label swaps on counterfactual groups. Overall, our results motivate assessing segmentation robustness under distribution shift beyond overlap by decomposing foreground errors into correct pixels, flipped-identity pixels, and missed-to-background pixels. We also propose an entropy-based, ground truth label-free `flip-risk' score, which is computed from foreground identity uncertainty, and show that it can flag flip-prone cases at inference time. Code is available at https://github.com/acharaakshit/label-flips.

阅读与讨论 → 访问原文 →

25.

arXiv (quant-ph) 2026-06-19 DOI: arXiv:2602.02466

Resolving problems with the continuum limit in coherent-state path integrals

作者:

Oliwier Urba\'nski↗

arXiv:2602.02466v2 Announce Type: replace Abstract: The paper solves the problem of continuum limit in bosonic thermal coherent-state path integrals. For this purpose, exact discrete versions of the path integral are constructed for three different orderings of the Hamiltonian: normal, anti-normal and symmetric (Weyl order). Subsequently, their different continuum versions are checked on the harmonic oscillator, to choose the symmetric ordering as a possibly correct choice for all polynomial Hamiltonians. Spotted mathematical subtleties in the simple case serve as a clue to the general solution. Finally, a general justification for the symmetric order is provided by deriving the continuum path integral starting from the exact discrete case using a renormalization procedure in the imaginary time frequency domain. While the role of Weyl order has already been found, the paper provides the missing proof of its suitability for every polynomial Hamiltonian and simplifies the previously established construction by referring only to creation and annihilation operators (without position and momentum operators).

阅读与讨论 → 访问原文 →