探索全球前沿学术脉络

01.

arXiv (CS.CL) 2026-06-12 DOI: arXiv:2606.13473

MaxProof: Scaling Mathematical Proof with Generative-Verifier RL and Population-Level Test-Time Scaling

作者:

Jiacheng Chen↗Xinyu Zhang↗Shunkai Zhang↗Yanmohan Wang↗Lin Li↗Tiancheng Qin↗Qin Wang↗Zhengmao Zhu↗Tianle Li↗Jingyang Li↗Zehan Li↗Binyang Jiang↗…

We present MaxProof, a population-level test-time scaling framework for competition-level mathematical proof in the MiniMax-M3 series. M3 first trains three proof-oriented capabilities – proof generation, proof verification, and critique-conditioned proof repair – using a defense-in-depth generative verifier engineered for low false-positive rate. These capabilities are merged into a single released M3 model. At test time, MaxProof treats the model as a generator, verifier, refiner, and ranker, searches over a population of candidate proofs, and returns one final proof through tournament selection. With MaxProof test-time scaling, the M3 model reaches 35/42 on IMO 2025 and 36/42 on USAMO 2026, exceeding the human gold-medal threshold on both.

阅读与讨论 → 访问原文 →

02.

arXiv (CS.AI) 2026-06-16 DOI: arXiv:2606.04990

From Agent Traces to Trust: A Survey of Evidence Tracing and Execution Provenance in LLM Agents

作者:

Yiqi Wang↗Jiaqi Zhang↗Taotao Cai↗Zirui Liu↗Qingqiang Sun↗Zequn Sun↗Zhangkai Wu↗Manqing Dong↗Mingkai Zhang↗Xuefei Yin↗Yanming Zhu↗

arXiv:2606.04990v2 Announce Type: replace-cross Abstract: Large language model (LLM)-based agents are evolving from passive text generators into autonomous systems capable of planning, tool use, retrieval, memory access, environmental interaction, and multi-agent collaboration. These capabilities expand agent autonomy, but also make agent behavior harder to verify, debug, and audit. Final-answer accuracy alone cannot explain how an output was produced, which evidence supported each claim, whether tool calls were justified, how memory influenced later decisions, or where failures originated. This survey examines evidence tracing and execution provenance as foundations for process-level accountability in trustworthy LLM agents. We define execution provenance as the typed graph of an agent execution and evidence tracing as its projection onto evidence-support relations. This perspective connects retrieval grounding, claim support, tool-use safety, memory lineage, observability, debugging, audit, and recovery within a unified framework. We introduce a taxonomy covering trace sources, evidence and execution units, provenance relations, tracing granularity and timing, representation forms, and trust functions. We then review key methodological directions, including provenance representation, evidence attribution, tool-use provenance, runtime guardrails, provenance-bearing memory, observability, and failure diagnosis. Finally, we discuss benchmarks, datasets, metrics, and open challenges for building provenance-aware, auditable, and recoverable agent systems.

阅读与讨论 → 访问原文 →

03.

arXiv (CS.AI) 2026-06-17 DOI: arXiv:2606.17123

LineageMark: Multi-user White-box Watermarking for Contribution Tracing in Model Derivation Chains

作者:

Bingxue Zhang↗Xiaofeng Xu↗Feida Zhu↗

arXiv:2606.17123v1 Announce Type: cross Abstract: In open large language model (LLM) ecosystems, models are frequently adapted across multiple domains and applications, forming multi-stage derivation chains. Consequently, tracking and verifying historical contributions is essential for model provenance and intellectual property protection. However, existing watermarking methods are mainly designed for single-user, one-time embeddings, often fail under repeated model derivation and incremental updates. To address this problem, we propose LineageMark, a multi-user white-box watermarking framework for model derivation chains. The framework encodes watermarks in model parameters using a projection-based approach. Stable carriers are first selected to reduce sensitivity to model changes, each watermark bit is then represented as a projection statistic over these carriers. Additional watermark insertions introduce only bounded perturbations in the projection space, and margin constraints are used to maintain signal integrity. We evaluate the effectiveness of LineageMark in multi-stage model derivation chains. Experimental results show that LineageMark preserves contributor watermarks across multi-stage derivation and supports incremental multi-user watermark insertion. Furthermore, it exhibits robustness against perturbations such as re-watermarking, fine-tuning, quantization, and pruning.

阅读与讨论 → 访问原文 →

04.

arXiv (CS.CL) 2026-06-16 DOI: arXiv:2606.15070

Stop When Further Reasoning Won't Help: Attention-State Adaptive Generation in Reasoning Models

作者:

Jiakai Li↗Ke Qin↗Rongzheng Wang↗Yizhuo Ma↗Qizhi Chen↗Muquan Li↗Shuang Liang↗

By incorporating test-time compute scaling, large reasoning models (LRMs) can solve complex problems through explicit chain-of-thought (CoT) reasoning processes. However, they often suffer from overthinking, resulting in redundant token outputs and degraded accuracy. Current methods to mitigate this issue remain limited: training-based approaches require substantial computational resources, while training-free methods rely on well-crafted prompts or unreliable confidence signals. In this work, we investigate early stopping from the perspective of attention distributions and propose a simple method, ASAG, which infers the model's reasoning state and adaptively adjusts the generation strategy. The proposed framework is training-free and plug-and-play, enabling seamless integration into existing LRMs. Extensive experiments on nine benchmarks demonstrate consistent improvements across mainstream LRMs with varying parameter scales, including the DeepSeek-R1-Distill and Qwen3 series. Specifically, ASAG improves average accuracy by 3.2% while reducing the number of generated tokens by nearly 40% across all reasoning tasks on Qwen3-8B.

阅读与讨论 → 访问原文 →

05.

arXiv (CS.AI) 2026-06-12 DOI: arXiv:2606.12439

Position: Generative Engine Optimization Creates Underexamined Risks, Governance Must Target Concentration, Disclosure, and Academic Blind Spots

作者:

Yizhu Wen↗Nan Zhang↗Haohan Yuan↗Xun Chen↗Haopeng Zhang↗Hanqing Guo↗

arXiv:2606.12439v1 Announce Type: cross Abstract: Large language model (LLM) answer engines are increasingly used for information seeking, shifting visibility from ranked lists to synthesized answers. This enables Generative Engine Optimization (GEO), which targets LLM answer engines' evidence pool and generation. We analyze the search engine optimization (SEO) to GEO transition to identify two risks: (i) concentrated influence from low contestability and system sensitivity, and (ii) undisclosed commercial influence embedded in evidence and reasoning. We then formalize a general GEO pipeline to locate where optimization acts and compare academic and industry practices, revealing a third risk: (iii) academic-industry blind spots driven by visibility and evaluation asymmetries between offline setups and deployed systems. This position argues the need for answer-level governance and measurement: stronger contestability, high-precision disclosure, black-box auditing of material influence, and deployment-aligned metrics for exposure persistence.

阅读与讨论 → 访问原文 →

06.

arXiv (CS.CV) 2026-06-11 DOI: arXiv:2606.11200

Detecting AI-Generated Content on Social Media with Multi-modal Language Models

作者:

Chenyang Yang↗Shen Yan↗Yibo Yang↗Litao Hu↗Yuchen Liu↗Yuan Zeng↗Hanchao Yu↗Yinan Zhu↗Sumedha Singla↗Brian Vanover↗Huijun Qian↗Zihao Wang↗…

Generative AI has enabled the creation of photorealistic images and videos that are increasingly disseminated on social media, often used for spam, misinformation, manipulation, and fraud. Existing AI-generated content (AIGC) detection methods face challenges including poor generalization to new generation models, reliance on single modalities, and lack of interpretable explanations. We present our pipeline that mitigates these issues by continuously curating diverse multi-modal social media data and training a compact vision-language model for detection and explanation. Our model achieves state-of-the-art detection performance on public benchmarks and demonstrates robust detection and explanation capabilities on internal social media datasets across multiple platforms. We deployed our model for post recommendation on social media platforms and observed positive downstream impacts on user engagement, demonstrating that it is feasible to perform effective AIGC detection in dynamic, real-world social media environments.

阅读与讨论 → 访问原文 →

07.

arXiv (CS.CL) 2026-06-12 DOI: arXiv:2606.13537

When Does Mixing Help? Analyzing Query Embedding Interpolation in Multilingual Dense Retrieval

作者:

Tongyao Zhu↗Chao-Ming Huang↗Min-Yen Kan↗

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.

阅读与讨论 → 访问原文 →

08.

arXiv (CS.CV) 2026-06-11 DOI: arXiv:2606.11740

UniReason-Med: A Shared Grounded Reasoning Interface for 2D-to-3D Transfer in Medical VQA

作者:

Mengzhuo Chen↗Yan Shu↗Chi Liu↗Hongming Piao↗Xidong Wang↗Derek Li↗Bryan Dai↗

We study whether grounded reasoning supervision from abundant 2D medical images can improve 3D medical VQA when both input types are aligned through a common reasoning interface. We introduce UniReason-Med, a single-checkpoint framework that processes either a 2D image or a slice-serialized 3D volume at inference time, generating interleaved textual reasoning and localized visual evidence through shared box syntax, region-token injection, and a common grounded reasoning policy. To train this interface, we construct UniMed-CoT, a 220K instruction-tuning dataset with interleaved textual reasoning and grounded visual evidence, including 170K 2D and 50K 3D samples. Through supervised fine-tuning followed by outcome-level reinforcement learning, UniReason-Med learns to generate grounded reasoning traces without IoU/Dice-based localization rewards during RL. Data-mixture and component ablations show that joint 2D+3D grounded supervision substantially improves 3D reasoning over 3D-only training, while grounding and region-token injection consistently benefit both 2D and 3D tasks. These results suggest that a shared grounded reasoning interface can transfer reasoning structure from 2D images to slice-serialized volumetric medical understanding. The code and data are publicly available at https://github.com/IQuestLab/unireason-med.

阅读与讨论 → 访问原文 →

09.

arXiv (CS.AI) 2026-06-19 DOI: arXiv:2606.19950

Confidence Calibration for Multimodal LLMs: An Empirical Study through Medical VQA

作者:

Yuetian Du↗Yucheng Wang↗Ming Kong↗Tian Liang↗Qiang Long↗Bingdi Chen↗Qiang Zhu↗

arXiv:2606.19950v1 Announce Type: cross Abstract: Multimodal Large Language Models (MLLMs) show great potential in medical tasks, but their elicited confidence often misaligns with actual accuracy, potentially leading to misdiagnosis or overlooking correct advice. This study presents the first comprehensive analysis of the relationship between accuracy and confidence in medical MLLMs. It proposes a novel method that combines Multi-Strategy Fusion-Based Interrogation (MS-FBI) with auxiliary expert LLM assessment, aiming to improve confidence calibration in Medical Visual Question Answering (VQA). Experiments demonstrate that our method reduces the Expected Calibration Error (ECE) by an average of 40\% across three Medical VQA datasets, significantly enhancing MLLMs' reliability. The findings highlight the importance of domain-specific calibration for MLLMs in healthcare, offering a more trustworthy solution for AI-assisted diagnosis.

阅读与讨论 → 访问原文 →

10.

arXiv (CS.CV) 2026-06-16 DOI: arXiv:2604.28185

Visual Generation in the New Era: An Evolution from Atomic Mapping to Agentic World Modeling

作者:

Keming Wu↗Zuhao Yang↗Kaichen Zhang↗Shizun Wang↗Haowei Zhu↗Sicong Leng↗Zhongyu Yang↗Qijie Wang↗Sudong Wang↗Ziting Wang↗Zili Wang↗Hui Zhang↗…

Recent visual generation models have made major progress in photorealism, typography, instruction following, and interactive editing, yet they still struggle with spatial reasoning, persistent state, long-horizon consistency, and causal understanding. We argue that the field should move beyond appearance synthesis toward intelligent visual generation: plausible visuals grounded in structure, dynamics, domain knowledge, and causal relations. To frame this shift, we introduce a five-level taxonomy: Atomic Generation, Conditional Generation, In-Context Generation, Agentic Generation, and World-Modeling Generation, progressing from passive renderers to interactive, agentic, world-aware generators. We analyze key technical drivers, including flow matching, unified understanding-and-generation models, improved visual representations, post-training, reward modeling, data curation, synthetic data distillation, and sampling acceleration. We further show that current evaluations often overestimate progress by emphasizing perceptual quality while missing structural, temporal, and causal failures. By combining benchmark review, in-the-wild stress tests, and expert-constrained case studies, this roadmap offers a capability-centered lens for understanding, evaluating, and advancing the next generation of intelligent visual generation systems.

阅读与讨论 → 访问原文 →

11.

arXiv (CS.CL) 2026-06-15 DOI: arXiv:2606.14150

Small LLMs: Pruning vs. Training from Scratch

作者:

Yufeng Xu↗Taiming Lu↗Kunjun Li↗Jiachen Zhu↗Mingjie Sun↗Zhuang Liu↗

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.

阅读与讨论 → 访问原文 →

12.

arXiv (CS.LG) 2026-06-16 DOI: arXiv:2606.15127

Beyond Accuracy: Measuring Bias Acknowledgment in Chain-of-Thought Reasoning for Responsible AI Evaluation

作者:

Xian Sun↗Wei Gao↗Yingshuo Wang↗Lingdong Kong↗Yanhang Li↗Zhichao Fan↗Zexin Zhuang↗Wenlong Dong↗Zhiyuan Zheng↗Hrishikesh Paranjape↗Abhishek Mandal↗Johnny R. Zhang↗…

arXiv:2606.15127v1 Announce Type: new Abstract: Reasoning models are increasingly used in settings where the final answer is not the only object of review: educational tools may show students intermediate steps, decision-support systems may require human oversight, and audit workflows may inspect traces for misleading or biased input. In such settings, two responses can receive the same final-answer score while differing in whether the trace explicitly flags injected biasing content. Accuracy-only evaluation collapses these cases. We study this gap as a measurement blind spot for responsible evaluation and introduce a minimal trace-level diagnostic with two axes: susceptibility (whether the bias breaks a previously correct answer) and acknowledgment (whether the trace contains a rubric-defined surface reference to the injected content). Across thousands of biased GSM8K trials, GPT-4o and Claude Sonnet~4 have similar susceptibility rates ($1.3\%$ vs.\ $1.2\%$) but substantially different acknowledgment rates ($13.0\%$ vs.\ $75.0\%$) under the same rubric.

阅读与讨论 → 访问原文 →

13.

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 →

14.

bioRxiv (Bioinfo) 2026-06-15 DOI: HASH:5dd8dc1ab90e4dcbdcb47162af2aa323

DAQplugin: Deep Learning based Real-time Model Evaluation Plugin for ChimeraX

作者:

Terashi↗Zhu↗Kihara↗

Although an increasing number of protein structures are determined by cryogenic electron microscopy (cryo-EM), protein structure modeling frequently suffers from residue misassignments and sequence register shifts, particularly in regions with ambiguous density. Here, we present DAQplugin, a ChimeraX plugin that performs real-time evaluation of protein models against cryo-EM density maps using the deep-learning-based residue-wise model quality (DAQ) score. Unlike existing validation tools that are typically applied after model construction, DAQplugin enables real-time deep-learning-based validation during model building and refinement. To our knowledge, DAQplugin is the first tool that provides real-time deep-learning based validation of protein models for cryo-EM map within an interactive modeling environment. In addition to identifying potential modeling errors, DAQplugin also provides guidance for correcting sequence register shifts by suggesting alternative residue placements along the backbone. The computation in this plugin is designed to run efficiently on general CPUs without requiring GPU hardware. Using DAQplugin, users can perform deep-learning-based validation on standard laptops during interactive model building, model-map fitting, and refinement. DAQplugin is able to facilitate more accurate interpretation of cryo-EM density maps and improve the reliability assessment of protein structure models.

阅读与讨论 → 访问原文 →

15.

arXiv (CS.LG) 2026-06-16 DOI: arXiv:2606.16304

pFedUL: Layer-Aware Federated Unlearning for Personalized Federated Learning

作者:

Zhuodong Liu↗Xiangyu Li↗Zhihao Zhang↗

arXiv:2606.16304v1 Announce Type: new Abstract: Federated unlearning (FU) enables the removal of specific data contributions from federated learning (FL) models to comply with regulations such as the General Data Protection Regulation (GDPR). However, most existing FU methods are designed for the FedAvg paradigm, where all clients share a single global model. In practice, personalized federated learning (pFL) methods such as FedPer, FedRep, Ditto, and FedBN have become widely adopted due to their superior handling of non-IID data. These methods decompose the model into shared global layers and client-specific personalized layers, fundamentally altering the semantics of unlearning, yet this setting has received little attention. We formalize FU under the pFL paradigm, identifying a tension between unlearning completeness on shared layers and personalization preservation for remaining clients. We then propose pFedUL, a layer-aware selective unlearning framework comprising three components: (1) gradient-based layer-wise contribution attribution that separately quantifies the target client's influence on shared and personalized parameters, (2) adaptive selective unlearning that applies differentiated forgetting strategies across layer types, and (3) a lightweight recalibration protocol enabling remaining clients to restore personalization with minimal overhead. We further introduce two new metrics, Personalization Preservation Score (PPS) and Cross-client Fairness Index (CFI), to evaluate pFL-specific unlearning quality. Experiments on CIFAR-10, CIFAR-100, and FEMNIST under varying non-IID settings indicate that pFedUL achieves unlearning effectiveness comparable to full retraining while maintaining an average of 97.3\% personalized accuracy for remaining clients. Compared with six state-of-the-art FU methods adapted to the pFL setting, pFedUL consistently achieves superior personalization preservation.

阅读与讨论 → 访问原文 →

16.

arXiv (CS.AI) 2026-06-16 DOI: arXiv:2606.15186

FreeSonic: Training-Free Temporal-Aware Decoupled Attention for Precise Audio Editing

作者:

Yuxuan Jiang↗Mingyang Han↗Yusheng Dai↗Andong Wang↗Tianhong Zhou↗Jiaxin Ye↗Dongxiao Wang↗Haoxiang Shi↗Boyu Li↗Jun Song↗Cheng Yu↗Bo Zheng↗…

arXiv:2606.15186v1 Announce Type: cross Abstract: Text-to-audio (TTA) generation has made significant strides, yet achieving precise and consistent audio editing remains a major challenge. However, existing methods struggle to balance temporal consistency with background preservation. In this paper, we propose FreeSonic, a training-free framework leveraging the state-of-the-art Rectified Flow-based TangoFlux model. FreeSonic utilizes an optimized inversion-reverse process and joint text-audio attention maps for precise target segment extraction. For content editing, a novel scheduled attention decoupling confines modifications to target regions while preserving original acoustic context. Furthermore, task-oriented noise injection enhances versatility for tasks such as audio removal and non-rigid replacement. Extensive experimental results demonstrate that FreeSonic achieves a superior balance by providing a high-fidelity and efficient solution for precise and consistent audio editing. Project and demos: https://free-sonic.github.io/

阅读与讨论 → 访问原文 →

17.

arXiv (CS.LG) 2026-06-11 DOI: arXiv:2606.11743

TacCoRL: Integrating Tactile Feedback into VLA via Simulation

作者:

Siyu Ma↗Yuqi Liang↗Chang Yu↗Yunuo Chen↗Hao Su↗Yixin Zhu↗Yin Yang↗Chenfanfu Jiang↗

arXiv:2606.11743v1 Announce Type: cross Abstract: Vision-language-action (VLA) models provide strong visual, language, and action priors for robot manipulation, but visual observations alone often miss the local contact state required for contact-rich tasks. We present TacCoRL, a scalable framework that injects Tactile feedback into VLA policies and improves them through sim-real Co-training and simulation-based reinforcement learning (RL), without requiring large-scale tactile pretraining or extensive real-world contact exploration. The key idea is not only adding touch as an input, but learning how contact readings should modulate action responses in near-failure states that are rare in demonstrations and risky to collect on hardware. We use a real-aligned simulator as a closed-loop training environment for contact interaction. Mixed simulated and real trajectories first warm-start tactile-conditioned actions in the pretrained policy. Reinforcement learning with verifiable task rewards then optimizes the policy using simulated contact rollouts. It reinforces tactile-conditioned actions that lead to task completion, while a supervised objective on real trajectories keeps the refined policy anchored to deployment visual, tactile, and action distributions. The resulting policy transfers directly to the real robot without privileged simulation state or online real-world RL. Across four bimanual contact-rich tasks, the final visuo-tactile policy achieves an average success rate of 72.5%, compared to baseline of 50.0%. Result videos and more details are available at https://tac-corl.github.io/

阅读与讨论 → 访问原文 →

18.

arXiv (CS.AI) 2026-06-17 DOI: arXiv:2606.17459

Can LLMs Be CEOs? Benchmarking Strategic Resource Reallocation with Multi-Role Agent Simulation

作者:

Yuyang Dai↗Xueqing Peng↗Lingfei Qian↗Zhuohan Xie↗

arXiv:2606.17459v1 Announce Type: new Abstract: Evaluating the decision-making capabilities of large language models (LLMs) is a growing research priority, yet existing benchmarks focus on isolated cognitive tasks such as reasoning, knowledge retrieval, and economic rationality in stylized settings. These evaluations overlook the defining challenge of real executive decision-making: integrating conflicting recommendations from specialized stakeholders under information asymmetry, organizational constraints, and temporal dependencies. We introduce \textsc{CEO-Bench}, a multi-agent benchmark that evaluates LLMs on CEO-level strategic resource reallocation – the process of redirecting capital across business units in a multi-round, constraint-rich organizational environment. In \textsc{CEO-Bench}, LLM agents receive conflicting advice from four role-conditioned C-suite advisors (CFO, CTO, COO, CMO), each with private signals and distinct priorities, and must synthesize these into a concrete allocation plan evaluated along four dimensions: role integration, conditional boldness, history-sensitive judgment, and plan validity. Experiments across five frontier models on 13 scenarios reveal that all models achieve high structural validity but diverge sharply on strategic calibration – the hardest capability layer. We identify systematic failure modes including single-advisor capture, conservative default under ambiguity, and historical amnesia, and uncover a structural integration-boldness tradeoff: models that engage more deeply with conflicting perspectives tend to produce less decisive action. These findings delineate the current capability boundary of LLMs as organizational decision-makers and inform the design of future AI-assisted executive systems.

阅读与讨论 → 访问原文 →

19.

arXiv (CS.CV) 2026-06-18 DOI: arXiv:2603.11417

Zero-Shot Cross-City Generalization in End-to-End Autonomous Driving: Self-Supervised versus Supervised Representations

作者:

Fatemeh Naeinian↗Ali Hamza↗Haoran Zhu↗Anna Choromanska↗

End-to-end autonomous driving models are typically trained on multi-city datasets using supervised ImageNet-pretrained backbones, yet their ability to generalize to unseen cities remains largely unexamined. When training and evaluation data are geographically mixed, models may implicitly rely on city-specific cues, masking failure modes that would occur under real-world domain shifts when generalizing to new locations. In this work, we formulate zero-shot cross-city transfer as a controlled representation-level stress test for end-to-end autonomous driving and ask how visual pretraining affects transfer behavior under geographic domain shift. We conduct a comprehensive study by integrating self-supervised backbones I-JEPA, DINOv2, and MAE into planning frameworks. We evaluate performance under strict geographic splits on nuScenes in the open-loop setting and on NAVSIM in the closed-loop evaluation protocol. Our experiments reveal a substantial generalization gap when transferring models across cities with different road topologies, traffic conventions, and visual environments. In open-loop evaluation, a supervised backbone exhibits severe degradation when transferring between cities, yet some domain-specific self-supervised methods can substantially reduce both displacement and collision degradation. In closed-loop evaluation, self-supervised pretraining improves average out-of-distribution PDMS in several single-city training settings. Our results provide empirical evidence that representation learning influences the robustness of cross-city planning and motivate zero-shot geographic transfer as an important stress test for evaluating end-to-end autonomous driving systems.

阅读与讨论 → 访问原文 →

20.

arXiv (quant-ph) 2026-06-15 DOI: arXiv:2606.06442

Nanostructure modelling with early fault tolerant quantum computers

作者:

Zhu Sun↗Christian Binder↗Balint Koczor↗Simon Benjamin↗

arXiv:2606.06442v2 Announce Type: replace Abstract: Semiconductor nanostructures are central to many developing technologies. Notably, double quantum dots are especially important for semiconductor spin-qubit architectures, quantum sensing applications, and quantum-dot solar cells. Accurate modelling is highly desirable but conventional methods can struggle when dynamics involve more than two interacting electrons. In this work, we present a quantum simulation framework capable of addressing multi-electron double quantum dots. We adopt an efficiently scaling 1$^st$ quantised representation of the system and develop algorithms based on both Trotterisation and Qubitisation. Incorporating insights from classical simulations enables us to produce resource estimates that are more realistic than those obtained from theoretical error bounds. Using a standard surface code model with physical noise at $10^{-3}$, our results indicate that the ground-state energy of four electrons in a double quantum dot can be estimated in approximately 22 hours using 226k physical qubits, or an eight-electron system in 3.3 days with 314k qubits (with runtimes falling dramatically when more qubits are available). We anticipate that incorporating recent advances in surface code architectures may reduce these costs significantly further. Our results suggest that early fault-tolerant quantum computers may become valuable tools for designing mature-era quantum technologies.

阅读与讨论 → 访问原文 →

21.

arXiv (CS.LG) 2026-06-15 DOI: arXiv:2601.14033

Private Prediction via PAC Privacy

作者:

Xiaochen Zhu↗Mayuri Sridhar↗Srinivas Devadas↗

arXiv:2601.14033v2 Announce Type: replace Abstract: Machine learning models are increasingly served behind APIs. This renders private prediction, i.e., privatizing a model's outputs rather than its parameters, a natural privacy target: model outputs are lower-dimensional and far more stable to training-data changes than weights. While differential privacy (DP) cannot effectively exploit this as it calibrates noise to worst-case sensitivity that is intractable to bound for non-convex models, we argue that PAC privacy is a natural fit for private prediction. It is instance-based, and calibrates noise to a black-box function's empirical stability to control mutual-information (MI) leakage. The missing ingredient is efficient, adaptive composition. Serving predictions means answering a long stream of adaptively chosen queries from untrusted users; existing composition either fails under adaptivity, grows quadratically, or reverts to input-independent, DP-like noise. We close this gap with a new adversarial composition result via adaptive noise calibration and prove that MI accumulates only linearly under adaptive and adversarial querying. Experiments across modalities show that prediction stability enables high utility even at a tiny per-query budget: on CIFAR-10, we achieve 87.79% accuracy with a per-query MI budget of $2^{-32}$. This enables serving one million queries while provably bounding membership-inference success to 51.08% – the same guarantee as $(0.04, 10^{-5})$-DP. Further, in the presence of auxiliary public data, the large volume of PAC-private predictions enables us to distill a publishable model that can be queried without limit. Concretely, 210,000 private labels on an ImageNet subset distill into a student reaching 91.86% accuracy on CIFAR-10 with membership inference success bounded by 50.49%, comparable to $(0.02, 10^{-5})$-DP.

阅读与讨论 → 访问原文 →

22.

arXiv (CS.AI) 2026-06-16 DOI: arXiv:2606.16262

UXBench: Measuring the Actionability of LLM-Generated UX Critiques

作者:

Wenjie Wang↗Yue Huang↗Zipeng Ling↗Han Bao↗Hang hua↗Xiaonan Luo↗Yu Jiang↗Shiyi Du↗Yuexing Hao↗Xiaomin Li↗Yuchen Ma↗Dianzhuo Wang↗…

arXiv:2606.16262v1 Announce Type: cross Abstract: Large language models (LLMs) are increasingly deployed as UX judges that inspect interfaces, diagnose usability problems, and propose repairs. Yet no controlled benchmark measures whether the resulting critiques are reliable and actionable across heterogeneous product surfaces. We introduce UXBench, a benchmark for evaluating LLMs as interaction-grounded UX judges. UXBench comprises local-first runnable web fixtures spanning ten product-surface families, paired with coverage-gated browser exploration that forces models to collect interaction evidence before reporting. Each judge model produces a structured UX report over seven rubric dimensions; report quality is measured by whether a fixed downstream repair agent can improve the interface based on the critique. We evaluate eight frontier models under both an automated repair-lift protocol and a blind human validation study. Results show that UX judging is neither saturated nor one dimensional: models differ meaningfully in report actionability, exhibit distinct rubric-level repair signatures, vary in fixture-level reliability, and trade leadership across surface categories

阅读与讨论 → 访问原文 →

23.

arXiv (CS.CV) 2026-06-17 DOI: arXiv:2605.15980

Flash-GRPO: Efficient Alignment for Video Diffusion via One-Step Policy Optimization

作者:

Xiaoxuan He↗Siming Fu↗Zeyue Xue↗Weijie Wang↗Ruizhe He↗Yuming Li↗Dacheng Yin↗Shuai Dong↗Haoyang Huang↗Hongfa Wang↗Nan Duan↗Bohan Zhuang↗…

Group Relative Policy Optimization has emerged as essential for aligning video diffusion models with human preferences, but faces a critical computational bottleneck: training a 14B parametered model typically demands hundreds of GPU days per experiment. Existing efficiency methods reduce costs through sliding window subsampling training timesteps, but fundamentally compromise optimization, exhibiting severe instability and failing to reach full trajectory performance. We present Flash-GRPO, a single-step training framework that outperforms full trajectory training in alignment quality under low computational budgets while substantially improving training efficiency. Flash-GRPO addresses two critical challenges: iso-temporal grouping eliminates timestep-confounded variance by enforcing prompt-wise temporal consistency, decoupling policy performance from timestep difficulty; temporal gradient rectification neutralizes the time-dependent scaling factor that causes vastly inconsistent gradient magnitudes across timesteps. Experiments on 1.3B to 14B parameter models validate Flash-GRPO's effectiveness, demonstrating substantial training acceleration with consistent stability and state-of-the-art alignment quality.

阅读与讨论 → 访问原文 →

24.

bioRxiv (Bioinfo) 2026-06-12 DOI: HASH:34f67b652901d7ce2ff7b64d2f180553

A Graph-based QSAR Modeling Pipeline for Predicting In vitro PubChem Assays and In vivo Human Hepatotoxicity: Mechanistic Analysis of Caspase-3/7 Activation

作者:

Chitikela↗Zhu↗Jia↗

Background: Caspase-3 and -7 are key effector caspases in the apoptotic pathway, a form of programmed cell death, and their activities serve as a well-established biomarker for evaluating environmental chemical toxicity and informing chemical risk assessment. Loss of mitochondrial membrane potential is a key event in the activation of Caspase-3/7 signaling and the subsequent induction of apoptosis. Therefore, simultaneous assessment of mitochondrial membrane potential and Caspase-3/7 activity enables elucidation of the mechanisms and pathways through which apoptosis is initiated. Rapid and accurate assessment of the potential toxicity of environmental chemicals and drugs remains a major challenge. Quantitative Structure Activity Relationship (QSAR) modeling have been widely used for toxicity prediction. Graph-based approaches encode compounds directly as molecular graphs, allowing structure-activity relationships to be learnt from molecular topology without the information loss in binary fingerprints. While advanced graph models such as graph transformers (GTs) have shown outstanding performance in many domains, they have not been fully leveraged in QSAR modeling on Caspase and mitochondrial toxicity. Methods: We propose a QSAR modeling pipeline that encompasses assay data preprocessing, feature representations (fingerprints and molecular graphs), and benchmarking machine learning (ML) models, including classic ML models, graph neural networks (GNNs), GTs, and their consensus ensembles. Based on in vitro Caspase and mitochondrial assays in PubChem, we applied the pipeline to predict Caspase-3/7 activation and mitochondrial membrane potential (MMP). Beyond in vitro assays, we also built in vivo QSAR modeling for FDA Drug-Induced Liver Injury (DILI) gold standard on human hepatotoxicity. Moreover, mechanistic analysis on Caspase-3/7 activation was conducted by comparing with MMP disruption to identify chemical substructures that may be responsible for dual activations. We also investigated cell-line-specific responses by identifying structural motifs that selectively induce Caspase-3/7 activation in individual cell lines.Results:Experimental evaluations show that GTs and GNNs outperformed classic ML models when the number of active compounds is large, such as MMP disruption, while classic ML models and GTs performed good for highly imbalance data with limited active compounds, such as Caspase-3/7 activation. For DILI prediction, the full consensus model achieved the highest AUC 0.69 and Graphormer had the highest F1 score 0.79, both surpassing the previous best model with AUC 0.63 and F1 0.65 with a large margin.Our mechanistic analysis shows that phenolic compounds bearing a para-hydroxyphenyl motif, as well as members of the lipophilic chain family with long alkyl chains can trigger the collapse of MMP, leading to the activation of caspases-3 and -7. Human embryonic kidney (HEK293) was the only cell line with a distinct structural motif: 1,1-dichloroethane and chlorobenzene. Human neuroblastoma (SK-N-SH) is uniquely impacted by an epoxide fragment and rat hepatoma (H-4-II-E) is uniquely impacted by a tetramethylcyclohexene motif and an acetaldehyde fragment.Conclusions:The proposed pipeline for QSAR modeling, including data preprocessing, feature representations, and incorporation of advanced graph ML approaches, is highly effective in predicting not only on Caspase-3/7 activation and membrane potential collapse, but also on FDA DILI human hetatotoxicity. As future research directions, we will leverage extra information, e.g., biological activity and findings in existing toxicity literature, and recent advances in large language models and agentic AI to further improve the predictive performance and enable a sensitive and specific framework for assessing human hepatotoxicity of environmental compounds.

阅读与讨论 → 访问原文 →

25.

arXiv (CS.CV) 2026-06-11 DOI: arXiv:2606.11670

ARGUS: Stacked Multi-View Identity Mosaic Injection for Subject-Preserving Video Generation

作者:

Zijie Meng↗Jiwen Liu↗Yufei Liu↗Chengzhuo Tong↗Xiaoqiang Liu↗Yuanxing Zhang↗Yulong Xu↗Pengfei Wan↗

Subject-preserving video generation is not solved by frontal-face similarity alone: a generated person must remain recognizable across motion, large viewpoint changes, expression shifts, occlusion, scale variation, and conflicts among text, first-frame, and identity references. We argue that the central bottleneck is the point-reference paradigm, which collapses identity into a single static observation entangled with pose, accessories, lighting, background, and camera statistics. We introduce Argus, a Wan-based framework centered on Stacked Multi-View Identity Mosaic Injection (SMII). SMII converts MLLM-selected image/video identity evidence into a 3*3 stacked mosaic, synchronizes the mosaic with the current diffusion time, and injects it as negative-time read-only memory in Wan's native token space. This turns identity from an external clean adapter or a single reference image into a compact dynamic distribution. Around SMII, an MLLM Identity Director selects informative identity moments and resolves condition conflicts, while no-cross-pair counterfactual training, Temporal Identity Annealing, and Adaptive Self-Likeness Guidance improve robustness without paired subject-video supervision. We further release HardID-Celeb, a public-figure identity-stress benchmark, and introduce YawScore and OccScore to probe large-yaw and first-frame-occlusion robustness. Argus achieves state-of-the-art results on OpenS2V-Eval Human-Domain, reaching 64.38 Total Score, 71.86 FaceSim, 51.62 NexusScore, and 79.14 NaturalScore. On HardID-Celeb, Argus obtains 76.80 FaceSim and improves YawScore and OccScore by 12.60 and 15.10 points over the strongest baselines, demonstrating that dynamic identity memory and large-scale counterfactual self-supervision are highly effective for subject-preserving video generation.

阅读与讨论 → 访问原文 →