探索全球前沿学术脉络

01.

bioRxiv (Bioinfo) 2026-06-17 DOI: HASH:3407e5cceceab8ab2518b7246c75e0f7

In silico characterization of lysis and host-recognition modules in Staphylococcus aureus bacteriophage genomes

作者:

Hasugian↗I. A↗Alifiyah↗N. I↗

Background/aim: Antimicrobial resistance in methicillin-resistant Staphylococcus aureus (MRSA) requires precision non-antibiotic therapeutics, yet phage lytic efficacy is poorly predicted by phenotypic assays, as shown by paradoxical biofilm responses. This study characterized the genomic architecture of lytic S. aureus bacteriophages, focusing on the conservation of the lysis module and the variability of host-recognition modules, to provide a rational basis for phage candidate selection. Materials and methods: Twenty-two complete S. aureus phage genomes were retrieved from NCBI GenBank. Genomic features were extracted with custom Biopython scripts. Lysis (endolysin, holin) and host-recognition (tail fiber/receptor-binding protein) modules were annotated and validated by InterPro domain analysis, with disrupted endolysins resolved by tBLASTn. Phylogeny was reconstructed from large terminase subunit (TerL) sequences using maximum likelihood. Results: Genome size spanned three classes, from 17.5 to 148.6 kb. The LysK-type endolysin (CHAP, Amidase, SH3b) was highly conserved, whereas tail fiber/RBP genes were detected in only 14 of 22 phages. Domain analysis reclassified two proteins annotated as endolysins as virion-associated peptidoglycan hydrolases, and identified two independent mechanisms, HNH endonuclease insertion and intron splitting, that interrupt lysis-module genes and confound automated annotation. Maximum likelihood analysis recovered a strongly supported, highly conserved core clade with EW and SA13 as divergent lineages. Conclusion: Lysis modules are conserved whereas host-recognition modules are variable, indicating that host recognition rather than the lytic enzyme is the principal determinant of host range and the more rational target for phage selection and engineering.

阅读与讨论 → 访问原文 →

02.

arXiv (CS.AI) 2026-06-16 DOI: arXiv:2606.15508

ToolMenuBench: Benchmarking Tool-Menu Filtering Strategies for Reliable and Efficient LLM Agents

作者:

Rahul Suresh Babu↗Laxmipriya Ganesh Iyer↗

arXiv:2606.15508v1 Announce Type: new Abstract: Tool-augmented large language model agents increasingly operate over large tool libraries, but existing evaluations often focus on whether a model can call a tool correctly rather than how the visible tool menu shapes reliability, efficiency, and safety-relevant risk exposure. We introduce ToolMenuBench, a benchmark for evaluating tool-menu construction in multi-step LLM agents. ToolMenuBench varies tool-menu size, distractor type, state-dependent task structure, and risk exposure, and reports both filter-level and downstream agent metrics, including visible-tool count, risky-tool exposure, task success, wrong-tool calls, premature actions, and token usage. In a controlled evaluation across seven model backends, three tool-menu sizes, six filtering methods, and seven evaluation settings, CMTF improves task success from 32.1% under all-tools exposure to 85.7%, while reducing average token usage by roughly 98%. Causal minimal tool filtering achieves the strongest overall tradeoff, reducing visible tools, wrong-tool calls, premature actions, and risky-tool exposure relative to unfiltered exposure, lexical filtering, state-aware filtering, and broader causal-path baselines. ToolMenuBench provides a reusable evaluation framework for studying the agent-interface problem: which tools should be visible, when they should be visible, and under what cost or risk constraints.

阅读与讨论 → 访问原文 →

03.

medRxiv (Medicine) 2026-06-15 DOI: HASH:dfaacf92e7c80e9ee1c4b62852689ed8

Repurposing cardiovascular disease risk models to predict incident and co-occurring cardiovascular, cardiometabolic and neurocognitive outcomes.

作者:

Quill↗Chaturvedi↗van Vugt↗Hingorani↗A. D↗Schmidt↗A. F↗

Background: Cardiovascular disease (CVD), cardiometabolic and neurocognitive conditions share risk factors and frequently co-occur. We evaluated whether four established CVD risk prediction models (QRISK3, PCE, SCORE2, SCORE2-OP) can be repurposed to predict 10-year risk of these conditions and their co-occurrence with CVD. Methods: The models were recalibrated using 20% of the UK Biobank (UKB) and evaluated in the remaining 80%. We performed external validation using data from Clinical Practice Research Datalink (CPRD) Aurum, assessing model discrimination (c-statistics) and calibration (intercept and slope). We used permuted feature importance to determine the influence of each individual predictor in the models. Results: Depending on the model, the c-statistics for incident CVD ranged from 0.71 to 0.74 in the UKB test set (16,137 events). Discrimination was equal to or higher than CVD when evaluated against non-traditional CVD outcomes: 0.74 to 0.77 for heart failure (3,471 events), 0.72 to 0.73 for atrial fibrillation (9,213 events), 0.73 to 0.75 for peripheral arterial disease (1,927 events) and 0.80 to 0.82 for abdominal aortic aneurysm (595 events). For the multimorbidity endpoints, model discrimination ranged from 0.74 for the composite of CVD and T2DM (SCORE2-OP) to 0.83 for the composite of CVD and dementia or Parkinson's disease (QRISK3). When considering the onset of any cardiovascular, cardiometabolic, or neurocognitive outcome discrimination ranged from 0.71 to 0.72. The repurposed models slightly underestimated the predicted risk in the CPRD compared to the UKB: average difference in calibration intercept was at most -0.64. After age and sex, smoking status and systolic blood pressure contributed most to model predictions. Conclusions: Repurposed CVD models can be used to identify 10-year risk of many CVD-related conditions and their multimorbidity. These may be used to support risk-based approaches to prevention and screening. The repurposed models have been made available at: https://repurposed-cvd-risk-models.shinyapps.io/cvd_cmd_dementia_app/ Keywords: Risk prediction; cardiovascular disease; cardiometabolic disease; dementia; disease prevention.

阅读与讨论 → 访问原文 →

04.

arXiv (CS.AI) 2026-06-16 DOI: arXiv:2605.02427

The Model Knows, the Decoder Finds: Future Value Guided Particle Power Sampling

作者:

Tu Nguyen↗Matthieu Zimmer↗Rasul Tutunov↗Xiaotong Ji↗Haitham Bou Ammar↗

arXiv:2605.02427v3 Announce Type: replace Abstract: A recurring pattern in "reasoning without training" is that base LLMs already assign non-trivial probability mass to correct multi-step solutions; the bottleneck is locating these modes efficiently at inference time. Power sampling provides a principled way to bias decoding toward such modes by targeting p_theta(x)^alpha with alpha > 1, but practical approximations must account for future-dependent correction factors that determine which prefixes remain promising. We introduce Auxiliary Particle Power Sampling (APPS), a blockwise particle algorithm for approximating the sequence-level power target with a bounded population of partial solutions. APPS propagates hypotheses in parallel using proposal-corrected power reweighting and refines their survival through future-value-guided selection at resampling boundaries. This redistributes finite compute across competing prefixes rather than committing to a single unfolding path, while providing a direct scaling knob in the particle count and predictable peak memory. We instantiate the future-value signal with short-horizon rollouts and also study an amortized variant that replaces rollouts with a lightweight learned selection head. AMore broadly, APPS improves the accuracy–runtime trade-off of training-free decoding, further supporting the view that inference-time power approximation can recover gains often attributed to post-training.

阅读与讨论 → 访问原文 →

05.

arXiv (CS.CV) 2026-06-11 DOI: arXiv:2606.11614

Information-Theoretic Decomposition for Multimodal Interaction Learning

作者:

Zequn Yang↗Yake Wei↗Haotian Ni↗Zhihao Xu↗Di Hu↗

Multimodal learning hinges on capturing redundant, unique, and synergistic information across modalities, which collectively constitute multimodal interactions. A critical yet underexplored challenge is that these implicit interactions vary dynamically across samples. In this work, we present the first systematic, information-theoretic analysis highlighting why learning these dynamic, sample-specific interactions is critical for effective multimodal learning. Our analysis further reveals deficits in conventional paradigms at learning these distinct interaction types: modality ensemble approaches struggle to capture synergy, while joint learning paradigms often under-utilize redundant information. This highlights the need for an approach that can adaptively learn from different interaction types on a per-sample basis. To this end, we propose Decomposition-based Multimodal Interaction Learning (DMIL), a novel paradigm that explicitly models and learns from sample-specific interactions. First, we design a variational decomposition architecture to isolate the constituent interaction components. Second, we employ a new learning strategy that leverages these explicit interaction components in a fine-tuning process to achieve comprehensive interaction learning. Extensive experiments across diverse tasks and architectures demonstrate that DMIL consistently achieves superior performance by adapting to holistic sample-specific interactions. Our framework is flexible and broadly applicable, establishing an interaction-centric paradigm for multimodal learning. The code is available at https://github.com/GeWu-Lab/DMIL.

阅读与讨论 → 访问原文 →

06.

arXiv (CS.AI) 2026-06-19 DOI: arXiv:2606.19791

Cross-Dataset, Age, and Gender Generalization: A Comprehensive Analysis of Fine-Tuning Strategies for Low-Resource Children's ASR

作者:

Paban Sapkota↗Hemant Kumar Kathania↗Mikko Kurimo↗Sudarsana Reddy Kadiri↗Shrikanth Narayanan↗

arXiv:2606.19791v1 Announce Type: cross Abstract: The challenge associated with recognizing dysarthric speech primarily arises from pronounced acoustic variability attributed to impaired articulatory precision. Past research has demonstrated improved recognition through the use of hybrid DNN/HMM sequence discriminative training. This paper presents a comprehensive investigation of various combinations of acoustic features tailored to different Acoustic Models, offering suitable feature selections for each. The incorporation of Pitch features notably improved recognition performance, especially for sentence recognition tasks involving dysarthric speech. Through a systematic examination of the TORGO database, we have demonstrated the potential to enhance the performance of the state-of-the-art Factorized Time Delay Neural Network (F-TDNN) model for recognizing dysarthric speech. Our methods, implemented with the F-TDNN model, resulted in a 4.65\% relative improvement in isolated word recognition and a 4.63\% relative improvement in sentence recognition for dysarthric speech, compared to previous research. This improvement effectively compensates for speech variability, attributable to our deliberate selection of the number of overlapping frames between consecutive training example chunks.

阅读与讨论 → 访问原文 →

07.

arXiv (CS.AI) 2026-06-15 DOI: arXiv:2606.14211

Closing the Reflection Gap: A Free Calibration Bonus for Agentic RL

作者:

Yinglun Zhu↗

arXiv:2606.14211v1 Announce Type: new Abstract: 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.

阅读与讨论 → 访问原文 →

08.

arXiv (CS.AI) 2026-06-16 DOI: arXiv:2606.15306

LatentGym: A Testbed For Cross-Task Experiential Learning With Controllable Latent Structure

作者:

Daksh Mittal↗Tommaso Castellani↗Thomson Yen↗Naimeng Ye↗Fangyu Wu↗Minghui Chen↗Tiffany Cai↗Emmanouil Koukoumidis↗William Zeng↗Hongseok Namkoong↗

arXiv:2606.15306v1 Announce Type: cross Abstract: We envision continually learning agentic systems that become more useful over time: as they encounter sequences of related tasks, they should infer the hidden structure shared across those tasks and use it to improve future decisions. This cross-task experiential learning capability is pivotal in domains such as personalization and interactive assistance, but existing training/evaluation frameworks do not provide shared, controllable latent structures and cannot measure whether or why agents improve. We introduce LatentGym: a controllable suite in which each environment is organized around a ground-truth latent variable governing the structure across tasks. Our construction yields metrics that separate exploration (whether the agent's actions gather information about the latent) from exploitation (whether the agent uses what it has gathered). We demonstrate our suite on empirical studies addressing three questions: how and why frontier models fail to adapt across related tasks; whether post-training on related task sequences improves general cross-task adaptation, and where those gains come from; and how design choices such as inter-task feedback shape training dynamics and generalization. Together, these results establish a controlled foundation for studying how LLM agents learn from experience across tasks, and for designing agents that adapt more reliably in sequential, personalized, and interactive settings.

阅读与讨论 → 访问原文 →

09.

arXiv (CS.CL) 2026-06-16 DOI: arXiv:2606.15833

When Correct Edges Cannot Be Verified: A Provenance Gap in Incomplete KGQA and a Provenance-Favoring Completion Policy

作者:

Yongqi Kang↗Yu Fu↗Yong Zhao↗

Incomplete Knowledge Graph Question Answering (IKGQA) requires completing missing edges to continue reasoning. A growing line of work verifies completed edges against retrieved text, treating textual support as a proxy for edge quality. We ask a question that, to our knowledge, has not been systematically tested: does textual verifiability actually track correctness? Exploiting the gold deleted triples provided by the standard random-deletion protocol, we measure both. The finding is counterintuitive: among gold-correct completed edges, 76-96% have no supporting passage even under exhaustive retrieval, robustly across deletion rates (20%/40%), datasets (CWQ/WebQSP), and relation types (structural, commonsense, long-tail). Most Freebase-style facts simply do not occur as head-tail co-mentions in text. Textual faithfulness therefore measures provenance, not correctness – separated by a paradigm-level gap no in-corpus retrieval closes. This reframes edge completion. Since most completed edges – correct or not – are causally redundant for the answer (95-97% of correct answers do not depend on any unsupported edge), the central question shifts from "is the edge correct?" to "admit or abstain under provenance uncertainty?" Within this framing we present TGComplete, a provenance-favoring admission policy that retrieves evidence at a reasoning breakpoint, verifies a candidate through a lightweight loop, and abstains when support is absent. Against the generate-to-complete baseline GoG, it attains higher edge precision against gold (15-21% vs 3-14%), with no statistically detectable EM loss and 3.1-7.4 times higher strict faithfulness of admitted edges – at the cost of lower recall. We position TGComplete not as uniformly better, but as a principled point on a precision/provenance-recall trade-off, appropriate when auditability matters.

阅读与讨论 → 访问原文 →

10.

arXiv (CS.CV) 2026-06-17 DOI: arXiv:2606.17846

Qwen-RobotManip Technical Report: Alignment Unlocks Scale for Robotic Manipulation Foundation Models

作者:

Haoqi Yuan↗Zhixuan Liang↗Anzhe Chen↗Ye Wang↗Haoyang Li↗Pei Lin↗Yiyang Huang↗Zixing Lei↗Tong Zhang↗Jiazhao Zhang↗Jie Zhang↗Jingyang Fan↗…

Foundation models in language and multimodality achieve strong generalization by aligning heterogeneous data under a unified formulation and training at scale. In this report, we investigate whether this scaling recipe can be applied to robotic manipulation to achieve genuine generalization. This is challenging because, unlike text, manipulation data is heterogeneous by nature, expensive to collect, and narrow in diversity, making alignment and scale simultaneously difficult. We present Qwen-RobotManip, a generalizable Vision-Language-Action foundation model built on Qwen-VL. Qwen-RobotManip introduces a unified alignment framework across the representation, motion, and behavioral dimensions of manipulation, making large-scale multi-source training coherent rather than conflicting. This alignment capability in turn enables Qwen-RobotManip to absorb manipulation data at a scale that prior training regimes could not sustain. A human-to-robot synthesis pipeline converts egocentric hand demonstrations into robot trajectories across 15 platforms, and a rigorous curation pipeline harmonizes heterogeneous datasets. Using only open-source datasets and human videos without proprietary data collection, Qwen-RobotManip constructs a ~38,100-hour pretraining corpus and exhibits emergent generalization capabilities, including zero-shot instruction following, robustness to perturbations, reactive error recovery, and cross-embodiment transfer. We find that standard benchmarks fail to capture pretraining quality and instead adopt OOD settings including RoboCasa365, LIBERO-Plus, EBench, RoboTwin-Clean2Rand, RoboTwin-IF, and RoboTwin-XE. Qwen-RobotManip substantially outperforms prior state-of-the-art models, including $\pi$0.5, across all OOD settings, ranks 1st in RoboChallenge with a 20% relative improvement, and is validated on real-robot platforms including AgileX ALOHA, Franka, UR, and ARX.

阅读与讨论 → 访问原文 →

11.

arXiv (CS.CV) 2026-06-16 DOI: arXiv:2606.15134

Beyond Scalar Distances: Semantic Attribute Gradients from Frozen MLLMs for Visual Embeddings

作者:

Shubhang Bhatnagar↗Dheeraj Baiju↗Narendra Ahuja↗

Vision encoders for retrieval are typically trained with class-label supervision: each training pair reduces to a scalar that uniformly pushes the embedding apart or pulls it together, as if every visual attribute either differed or matched. A multimodal large language model (MLLM), shown the same pair, can articulate those attributes and use them to predict whether the images share a class. We propose SAGA, a framework that turns this language-grounded, attribute-aware perception into a training signal for the encoder itself. Specifically, we use Group Relative Policy Optimization (GRPO) to reward the MLLM for correct predictions on the vision encoder's tokens. Since correct predictions require those tokens to expose the specific attributes that differ or match between the pair, the gradient pushes the encoder to encode them, replacing the uniform pair-level scalar with attribute-resolved supervision. An auxiliary attention-distillation loss anchors the encoder's embedding to tokens the MLLM attended to, and a standard metric-learning loss shapes the embedding geometry for nearest-neighbour retrieval. The MLLM is frozen throughout and discarded at inference, matching the deployment cost of a metric-learning baseline. SAGA improves Recall@1 by 3 to 6 points over state-of-the-art baselines on CUB-200-2011, Cars-196, FGVC-Aircraft, and iNaturalist Aves on zero-shot image retrieval.

阅读与讨论 → 访问原文 →

12.

arXiv (quant-ph) 2026-06-19 DOI: arXiv:2606.20328

Effective Faraday interaction between light and Helium-3 nuclear spins in a multi-pass cell

作者:

Kaiwen Yi↗Yida Sha↗Zejia Lin↗Matteo Fadel↗Xiang Peng↗

arXiv:2606.20328v1 Announce Type: new Abstract: Helium-3 nuclear spins form an exceptionally stable quantum system with extremely long coherence time, offering exciting opportunities for quantum technologies. In particular, nuclear spin-squeezed states promise enhanced precision for sensing tasks and tests of new physics. A central challenge for all these applications is the realization of a controllable light-nuclear spin interface. Here we experimentally demonstrate such an interface by exploiting metastability-exchange collisions in a low-pressure helium-3 gas cell at room temperature. A radio-frequency discharge produces a small population of metastable atoms that both enables efficient optical pumping and mediates an effective Faraday interaction between the collective nuclear spin and an optical probe. We quantitatively characterize the strength of this interaction as a function of the nuclear polarization, applied magnetic field, and probe-beam parameters. Moreover, we show that using a multi-pass cell enhances this interaction by effectively increasing the optical depth. Extrapolating to a tenfold increase of the probe power used in the present experiment, we project a measurement-induced squeezing rate of 0.52 s$^{-1}$. Our results provide a practical pathway for optical access to helium-3 nuclear spins and open prospects for generating long-lived, macroscopic nuclear spin-squeezed states for quantum metrology.

阅读与讨论 → 访问原文 →

13.

arXiv (CS.CL) 2026-06-16 DOI: arXiv:2606.16383

Surpassing Scale by Efficiency: A Compact 135M Parameter Foundational LLM Natively Adapted for the Bangla Language

作者:

Rabindra Nath Nandi↗

While the NLP landscape is dominated by multi-billion parameter architectures, their deployment in low-resource, non-Latin scripts remains computationally prohibitive for edge configurations, mobile systems, and decentralized local hardware. This paper presents bangla-smollm-135m, a highly compact 135-million parameter decoder-only foundational model engineered explicitly for high-efficiency language modeling in the Bangla script. By leveraging a deterministic intersect-and-append token merging strategy between TituLLMs and SmolLM2-135M, the model overcomes subword script fragmentation without destabilizing early pretrained parameter states. In zero-shot multi-task benchmark evaluations (PIQA_bn, OpenBookQA_bn, CommonsenseQA_bn, and Bangla_MMLU), bangla-smollm-135m matches or outperforms models twice its size (Gemma-3-270m) and achieves parity with models in the 1B parameter tier. The model is available at rnnandi/bangla-smollm-135m

阅读与讨论 → 访问原文 →

14.

PLOS Medicine 2026-06-12 DOI: HASH:5a2bb8e07f5eec7c849273586d4bf467

Comparison of count-based and clustering definitions of multimorbidity and their association with prevalence of multimorbidity, health profiles, and mortality: A cohort study of UK Biobank participants

作者:

Gabriella C. Silva↗

by Gabriella C. Silva, Aurore Fayosse, Louis Jacob, Séverine Sabia, Archana Singh-Manoux, Benjamin Landré Background Multimorbidity, the presence of several chronic conditions, is linked to higher mortality and healthcare use and thus poses a major challenge for aging populations. While most studies rely on simple counts of conditions, clustering approaches have been proposed to describe patterns of co-occurring diseases. We aimed to evaluate the extent to which these methodological choices influence prevalence and association with health profiles and mortality. Methods and findings Using UK Biobank baseline data (n = 474,397), collected between 2006 and 2010, we compared six count-based definitions of multimorbidity based on different condition lists (extended, most prevalent, or body systems) and thresholds (≥2 versus ≥3 conditions). We also applied a clustering analysis to characterize subtypes of multimorbidity among participants with at least two chronic conditions. We compared prevalence and associations with concurrent health outcomes (polypharmacy, self-rated health, frailty, falls, surgery, chronic pain), blood-based measures (C-reactive protein, Cystatin-C, HDL, LDL Cholesterol, IGF-1), and 3- and 10-year mortality risks. Analyses were undertaken separately in men and women using multivariable regression models adjusted for sociodemographic characteristics and body mass index. Multimorbidity prevalence ranged from 1.0% (cluster-based) to 35.3% (count-based). Count-based definitions using lists with more conditions yielded higher prevalence. Higher thresholds identified more severe health profiles on all measured health outcomes, blood-based measures, but not higher mortality risks. Associations with blood-based measures were more pronounced using clustering, with the highest differences from the standard definition distributed across clusters. Odds ratios for 3-year mortality ranged from 1.44 [1.26; 1.64] to 4.60 [3.73; 5.62] for men and 1.35 [1.07; 1.69] to 3.83 [2.78; 5.14] for women. For 10-year mortality, they ranged from 1.42 [1.34; 1.50] to 3.86 [3.46; 4.30] in men and 1.29 [1.21; 1.39] to 3.33 [2.93; 3.77] for women, with clustering identifying groups with low prevalence and high mortality risks. Findings should be interpreted in light of the selected nature of the UK Biobank cohort and the cross-sectional assessment of several health indicators. Conclusion Operational definitions of multimorbidity substantially influence prevalence estimates, while associations with mortality appear more robust across count-based approaches. Clustering analyses provide complementary insights into heterogeneity within multimorbid populations. Future translational studies are warranted to determine how multimorbidity definitions can be optimized to ultimately improve clinical management and health outcomes in practice.

阅读与讨论 → 访问原文 →

15.

arXiv (CS.CL) 2026-06-12 DOI: arXiv:2606.12608

Shopping Reasoning Bench: An Expert-Authored Benchmark for Multi-Turn Conversational Shopping Assistants

作者:

Shuxian Fan↗Seonwoo Min↗Youna Hu↗Botao Xia↗Jayakrishnan Unnikrishnan↗Rowan Musselmann↗Yifan Gao↗Qingyu Yin↗Priyanka Nigam↗Bing Yin↗

Conversational shopping assistants now serve hundreds of millions of customers, yet no existing benchmark jointly evaluates the open-ended multi-turn reasoning, domain expertise, and criterion-level quality that real shopping conversations demand. Shopping reasoning is unique among language model applications. Unlike factual question answering or verifiable code generation, it requires balancing subjective preferences, budget constraints, and cross-product trade-offs across multi-turn dialogue, capabilities absent from previous e-commerce and general-purpose benchmarks. We introduce the Shopping Reasoning Bench, an expert-authored benchmark of 525 missions (232 single-turn, 293 multi-turn) with 10863 importance-weighted binary rubrics authored by retail domain experts. These criteria are organized under a taxonomy of five reasoning categories and fifteen subcategories covering diverse demands such as preference refinement, trade-off analysis, and compatibility assessment. An evaluation of nine models across three families (GPT, Claude, Gemini) shows that pass rates reach only 57–77% overall. On multi-turn missions, all models score 13–29 points lower on optional above-and-beyond criteria than on required ones, and performance degrades 4–18 points as conversations progress. These gaps show that current models handle basic shopping assistance but fall short of expert-level advice, making Shopping Reasoning Bench a challenging testbed for future shopping assistant development.

阅读与讨论 → 访问原文 →

16.

arXiv (CS.CL) 2026-06-18 DOI: arXiv:2606.18699

TW-LegalBench: Measuring Taiwanese Legal Understanding

作者:

Fei-Yueh Chen↗Chun Huang Lin↗Chan Wei Hsu↗Kuan Hsuan Yeh↗Zih-Ching Chen↗Kuan-Ming Chen↗Patrick Chung-Chia Huang↗

Large language models (LLMs) have shown impressive capabilities across diverse tasks, yet their performance on jurisdiction-specific legal reasoning remains underexplored. We present TW-LegalBench that utilizes Taiwanese legal system's rich official corpus open to the public to fill the gap in evaluating LLMs on Taiwanese law, among common-law benchmarks that focus on English sources and civil-law benchmarks focusing on sources of Simplified Chinese. TW-LegalBench comprises three task types: (1) over 16,000 multiple-choice questions (MCQs) across five years of official examinations in 18 professional domains; (2) 117 open-ended essay questions (OEQs) from examinations for legal professionals with official scoring rubrics; and (3) more than 14,000 legal judgment prediction (LJP) instances covering hundreds of crime categories. We evaluate 13 LLMs using accuracy for MCQs, a decomposed LLM-as-Judge framework based on the scoring rubric points for OEQs, and metrics for sentencing accuracy and statute citation for LJP. Our results reveal that top-performing models exceed the passing threshold for qualified lawyers (passing rate: 11%) but fall short of that for judges and prosecutors (passing rate: 1~2%). For LJP, while models demonstrate reasonable verdict type accuracy and sentence prediction capability, they struggle to cite exact legal articles. These findings highlight that reliable legal text generation remains challenging for LLMs, even though their performance on qualification examinations approaches human level.

阅读与讨论 → 访问原文 →

17.

arXiv (CS.CL) 2026-06-16 DOI: arXiv:2606.16583

Uncertainty Is Not a Safety Net for Clinical VQA, but Can It Anticipate Model Failure?

作者:

Arnisa Fazla↗Alberto Testoni↗Ameen Abu-Hanna↗Barbara Plank↗Iacer Calixto↗

Safe deployment of clinical vision-language models (VLMs) requires reliable uncertainty estimation (UE): a signal indicating when predictions should be trusted or escalated to a clinician. We test whether current UE methods actually deliver this signal. Benchmarking 8 methods across 12 VLMs on clinical visual question-answering (VQA), we find that UE quality is not an intrinsic property of the UE method: it tracks model accuracy, degrading precisely where the model performance is weakest, and therefore where reliability is most needed. When we stress-test models by hiding the correct option among the multiple-choice answers (NOTA perturbations), accuracy collapses while uncertainty barely changes, leaving models systematically miscalibrated. Yet, we find that uncertainty on the unperturbed input reliably anticipates which predictions will collapse under NOTA, indicating that UE in current VLMs carries diagnostic information about model fragility. Our results position UE as a diagnostic tool for identifying fragile predictions and motivate perturbation-based evaluation as a path toward safe clinical deployment.

阅读与讨论 → 访问原文 →

18.

arXiv (CS.CV) 2026-06-16 DOI: arXiv:2606.15861

Object Tokens as a Bridge Between Segmentation and Visual Question Answering in Robotic Surgery

作者:

Yiping Li↗Ronald de Jong↗Romy van Jaarsveld↗Franco Badaloni↗Gino Kuiper↗Jelle Ruurda↗Josien Pluim↗Marcel Breeuwer↗

Visual Question Answering (VQA) in robotic surgery, referred to as surgical VQA, requires high-level understanding of complex surgical scenes and the integration of visual perception with language reasoning, with the potential to support surgical training and intraoperative decision-making. Recent Vision-Language Models (VLMs) have shown promising performance through parameter-efficient fine-tuning; however, most existing approaches rely on coarse visual grounding, typically limited to bounding boxes, which fails to capture the fine-grained spatial structure of surgical objects. In this work, we propose a unified framework that jointly performs pixel-level segmentation and visual question answering within a single framework. Our approach integrates a VLM with a Segment Anything Model (SAM)-based decoder and represents scene elements as object tokens generated by the VLM. These object tokens guide answer prediction and are further projected to the SAM-based decoder to produce segmentation masks. By optimizing the object token embeddings through both segmentation and question answering objectives, the model learns spatially grounded representations that enhance visual reasoning while providing explicit pixel-level grounding. We evaluate the proposed method on the private RAMIE (Robot-Assisted Minimally Invasive Esophagectomy) dataset and the public EndoVis18 dataset, where it consistently outperforms baseline methods for surgical VQA. These results demonstrate that incorporating context-aware object tokens into vision-language models improves fine-grained surgical scene understanding.

阅读与讨论 → 访问原文 →

19.

arXiv (CS.CL) 2026-06-18 DOI: arXiv:2606.18502

Towards Scalable Customization and Deployment of Multi-Agent Systems for Enterprise Applications

作者:

Paresh Dashore↗Shreyas Kulkarni↗Uttam Gurram↗Nadia Bathaee↗Kartik Balasubramaniam↗Genta Indra Winata↗Sambit Sahu↗Shi-Xiong Zhang↗

Large language model (LLM)-based multi-agent systems demonstrate strong performance on complex reasoning and task execution, enabling broad enterprise applications. However, production deployment remains challenging due to domain-specific customization requirements and high latency and inference costs in agentic workflows. We propose a unified framework for customization and efficient deployment of multi-agent systems in real-world settings. The first stage, Agentic Model Customization, combines continual pretraining, supervised fine-tuning, and preference optimization to adapt a compact model to specialized domains while retaining strong agentic capabilities. The second stage, Inference Optimization, integrates speculative decoding and FP8 quantization with targeted calibration to enable cost-efficient serving with minimal quality loss. Across enterprise workloads, our framework enables rapid domain adaptation and achieves a 4.48x speedup in throughput while maintaining performance and improving robustness on long-tail scenarios.

阅读与讨论 → 访问原文 →

20.

arXiv (CS.CL) 2026-06-18 DOI: arXiv:2606.18473

PreUnlearn: Auditing Collateral Knowledge Damage Before Large Language Model Unlearning

作者:

Bo Su↗Ankit Shah↗Thai Le↗

Machine unlearning for large language models (LLMs) aims to remove specified knowledge while preserving the rest of the model's capabilities. However, the boundary between knowledge to forget and knowledge to retain is often unclear, since related and even distant information may be entangled in the model. In this paper, we study LLM unlearning from a data-centric perspective and measure how unlearning effects propagate from the forget set to same-domain and distant-domain knowledge. We find a consistent decay pattern: collateral damage is strongest near the forget set, weakens with semantic distance, but does not disappear at domain boundaries. We further ask whether such damage can be audited before unlearning is executed. We formulate forget-set auditing as a pre-unlearning prediction task and analyze which data features are most predictive of downstream damage. Our results show that interaction features between the forget set and evaluation set provide the strongest signals, suggesting that collateral damage is partly reflected in data geometry before model updates occur. These findings position forget-set auditing as an early warning tool for identifying risky unlearning runs and designing more reliable unlearning procedures.

阅读与讨论 → 访问原文 →

21.

arXiv (CS.CL) 2026-06-16 DOI: arXiv:2511.08577

Think-at-Hard: Selective Latent Iterations to Improve Reasoning Language Models

作者:

Tianyu Fu↗Yichen You↗Zekai Chen↗Guohao Dai↗Huazhong Yang↗Yu Wang↗

Improving the reasoning abilities of Large Language Models (LLMs), especially under parameter constraints, is crucial for real-world applications. Looped transformers address this by performing multiple latent iterations to refine each token beyond a single forward pass. However, we identify a latent overthinking phenomenon: most token predictions are already correct after the first pass, but are sometimes revised into errors in later iterations. We ask whether selectively skipping latent iterations can improve accuracy, and reveal significant potential with an oracle iteration policy that boosts performance by up to 7.3%. Motivated by this, we propose Think-at-Hard (TaH), a looped transformer optimized for selective iteration. TaH employs a lightweight neural decider to trigger latent iteration, only at tokens likely to be incorrect after the standard forward pass. During latent iterations, depth-aware Low-Rank Adaptation (LoRA) modules shift the objective from general next-token prediction to focused hard-token refinement. A duo-causal attention mechanism extends attention from the token sequence dimension to an additional iteration depth dimension, enabling cross-iteration information flow with full sequential parallelism. Experiments on nine benchmarks show consistent gains across math, QA, and coding tasks. With identical parameter counts, TaH outperforms always-iterate baselines by 3.8-4.4% while skipping iterations on 93% of tokens, and exceeds single-iteration Qwen3 baselines by 3.0-3.8%. When allowing

阅读与讨论 → 访问原文 →

22.

arXiv (quant-ph) 2026-06-16 DOI: arXiv:2606.16451

Detecting basis-dependent hardware errors through spatio-temporal quantum steering

作者:

Hsiang-Wei Huang↗Kuo-Feng Chiu↗Yi-Te Huang↗Jhen-Dong Lin↗Tse-Ming Chen↗Yueh-Nan Chen↗

arXiv:2606.16451v1 Announce Type: new Abstract: Spatio-temporal quantum steering provides a framework for benchmarking the nonclassicality of general quantum state transfer processes. A central diagnostic is the no-signaling-in-time (NSIT) condition, whose violation can indicate basis-dependent hardware errors. However, finite measurement statistics may also yield apparent violations, thereby obscuring the detection of basis-dependent hardware errors. To address this, we construct a statistical hypothesis test under the null hypothesis that NSIT violations arise solely from statistical fluctuations. Combining the statistical properties of NSIT violation under the null hypothesis with Chebyshev's inequality, we obtain a distribution-free upper bound on the $p$-value without parametric assumptions. We apply this method to two examples. For a single-qubit state-transfer experiment on a superconducting processor, we observe several instances that the NSIT violation is observed and the null hypothesis is simultaneously rejected by a small $p$-value, providing statistical evidence of basis-dependent hardware errors. For a seven-qubit Hayden-Preskill teleportation protocol on IonQ devices, the null hypothesis is also rejected even when the average fidelity exceeds the classical threshold, while the associated nonclassicality measure vanishes. Our results highlight the necessity of statistical hypothesis testing for detecting basis-dependent errors in near-term quantum devices.

阅读与讨论 → 访问原文 →

23.

arXiv (CS.AI) 2026-06-18 DOI: arXiv:2606.18950

RTSGameBench: An RTS Benchmark for Strategic Reasoning by Vision-Language Models

作者:

San Kim↗Daechul Ahn↗Reokyoung Kim↗Hyeonbeom Choi↗Seungyeon Jwa↗Jonghyun Choi↗

arXiv:2606.18950v1 Announce Type: new Abstract: Modern Vision-Language Models (VLMs) often struggle with strategic reasoning, i.e., anticipating and influencing other agents' actions, under uncertainty in competitive and cooperative settings. Real-time strategy (RTS) games can be a natural testbed for diagnosing this limitation, as they demand coordination with allies, adaptation to opponents' strategy, and long-horizon planning under partial observability. However, existing RTS benchmarks offer limited evaluation scope, lack systematic competency diagnosis, and remain fixed in the pre-designed scenario coverage. To address these limitations, we present RTSGameBench, which is built on Beyond All Reason, a large-scale RTS game with an expanded battlefield that demands broader strategy diversity than the existing testbeds. The proposed benchmark provides evaluations through diverse gameplay across various matchup structures, diagnostic assessment via mini-games, each targeting an individual strategic competency, and extensible coverage via a self-evolving generation framework that converts free-form queries into new mini-games, improving over successive cycles. Additionally, for VLMs to operate in large-scale RTS games, we provide RTSGameAgent that manages units by an FSM with agentic memory. We empirically validate that multiple state-of-the-art VLMs do not perform well when matchups demand tighter coordination, multiagent coordination and when task scale increases.

阅读与讨论 → 访问原文 →

24.

arXiv (CS.AI) 2026-06-12 DOI: arXiv:2606.13392

MiniMax Sparse Attention

作者:

Xunhao Lai↗Weiqi Xu↗Yufeng Yang↗Qiaorui Chen↗Yang Xu↗Lunbin Zeng↗Xiaolong Li↗Haohai Sun↗Haichao Zhu↗Vito Zhang↗Pengyu Zhao↗

arXiv:2606.13392v1 Announce Type: new Abstract: 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.

阅读与讨论 → 访问原文 →

25.

arXiv (quant-ph) 2026-06-19 DOI: arXiv:2606.20020

Effects of interaction range on the mean-field dynamics of Bose polarons

作者:

Piotr Wysocki↗Ubaldo Cavazos Olivas↗Marek Tylutki↗Krzysztof Jachymski↗

arXiv:2606.20020v1 Announce Type: cross Abstract: We consider the three-dimensional Bose polaron problem in the regime of finite range interactions and competing length scales. Working in the reference frame of the impurity, we study both static and out of equilibrium properties of the system, in particular the transfer of momentum between the impurity and the host gas. We find that relaxation dynamics can occur via damped oscillations of the impurity velocity with simple dependence on the interaction strength. Furthermore, the equilibration process is sensitive to the type of the impurity-bath interaction. Specifically, interatomic forces describing ion-atom systems lead to much longer timescales and more pronounced oscillations in the strong coupling regime with respect to local interaction potentials. We also find that the effective masses can differ by a large amount between the two scenarios, even if the number of atoms in the polaron cloud remains similar for both cases.

阅读与讨论 → 访问原文 →