本篇博文主要内容为 2026-07-27 从Arxiv.org论文网站获取的最新论文列表，自动更新，按照NLP、CV、ML、AI、IR、MA六个大方向区分。
说明：每日论文数据从Arxiv.org获取，每天早上12:30左右定时自动更新。
提示: 当天未及时更新，有可能是Arxiv当日未有新的论文发布，也有可能是脚本出错。尽可能会在当天修复。
目录
概览 (2026-07-27)
今日共更新466篇论文,其中:
- 自然语言处理共47篇(Computation and Language (cs.CL))
- 人工智能共146篇(Artificial Intelligence (cs.AI))
- 计算机视觉共68篇(Computer Vision and Pattern Recognition (cs.CV))
- 机器学习共138篇(Machine Learning (cs.LG))
- 多智能体系统共7篇(Multiagent Systems (cs.MA))
- 信息检索共8篇(Information Retrieval (cs.IR))
- 人机交互共23篇(Human-Computer Interaction (cs.HC))
多智能体系统
[MA-0] RACE-ROUTER: Task-Consistent and Adaptive Online Routing for Agent ic AI
【速读】：该论文旨在解决企业在部署生成式 AI (Generative AI) 时，现有路由系统在处理长时序代理任务（agentic applications）中的根本性缺陷。传统路由机制对每次 LLM 调用独立决策，无法将最终任务级结果的反馈正确归因于单个路由选择，导致策略优化失效。其解决方案的关键在于提出一种基于任务级别的路由框架——TRACE-Router，该框架通过上下文相关强化学习（contextual bandit）在任务初始接入时一次性选定后端模型，并在整个任务执行过程中固定使用该模型，从而实现路由与监督单位的一致性。同时，利用任务完成后的延迟奖励信号（terminal reward），联合优化准确率与延迟，动态调整路由策略，无需显式估计任务复杂度。实验表明，在三个代理基准测试中，TRACE-Router 均实现了更优的精度-延迟权衡，达到非劣帕累托前沿点，在 tau2-Bench 上较延迟匹配的模型插值方案提升 7–8 个精度点，在 Terminal-Bench 上以 36% 更低延迟实现比最强单模型基线高 7.1 个精度点的性能。
链接: https://arxiv.org/abs/2607.22465
作者: Ritik Raj,Souvik Kundu,Sarbartha Banerjee,Dheemanth Joshi,Ishita Vohra,Tushar Krishna
机构: Georgia Institute of Technology(佐治亚理工学院); Intel(英特尔); Texas AM University(德克萨斯农工大学)
类目: Artificial Intelligence (cs.AI); Machine Learning (cs.LG); Multiagent Systems (cs.MA)
备注:
Abstract:Routing to select large language models (LLMs) with different cost-quality trade-offs has become a fundamental deployment feature of enterprise AI. Existing routers, primarily make independent routing decisions for each LLM call. However, agentic applications execute as long-horizon workflows whose quality is determined only by a delayed, task-level outcome. This mismatch prevents per-call routers from correctly attributing feedback to individual routing decisions. Towards mitigating this, we present TRACE-Router, a task-level routing framework that aligns routing with the unit of supervision. TRACE-Router assigns each task to a model once at admission using a contextual bandit, pins all subsequent LLM calls to the selected backend, and updates its policy using the task’s terminal reward, jointly accounting for accuracy and latency. By leveraging delayed task feedback, TRACE-Router learns routing policies that adapt to the workload while avoiding explicit task-complexity estimation. Across three agentic benchmarks, TRACE-Router consistently improves the accuracy-latency trade-off, achieving non-dominated Pareto frontier points. On tau2-Bench, it outperforms latency-matched interpolation between individual models by 7-8 accuracy points, while on Terminal-Bench it achieves 7.1 higher accuracy points than the strongest single model baseline with 36% lower latency.
[MA-1] Where FactsGo Missing: A LayerwiseTaxonomy and Per-Layer Attribution of Information Omissionin Air-Gapped LLM Agent Pipelines
【速读】：该论文旨在解决在受监管环境（如临床FHIR服务、法律审查、主权基础设施）中，由于隔离部署（air-gapped）和本地化运行（on-premises）限制，无法调用前沿大模型API时，所面临的生成式AI系统中关键信息遗漏（omission）这一严重可靠性问题。其核心挑战在于：系统在处理海量数据时，可能沉默地遗漏决策所需的关键事实（例如，仅分析400条记录中的20条却报告“无异常”），而此类遗漏并非由模型本身能力不足导致，而是贯穿整个推理流水线的系统性现象。解决方案的关键在于提出一个九层分类体系（L0-L8），将遗漏机制从数据摄入到智能体循环的全链路进行精确归因，并通过控制消融与对数概率分解方法，区分确定性中间件层（L0-L3）与行为性执行层（L4-L8），构建可量化的“遗漏瀑布”分析框架。研究进一步开发了一个开放的跨架构测试平台，对比滑动窗口混合、全注意力及状态空间模型（SSM）混合等不同架构在多种引擎与框架下的表现，并设计了一套适用于隔离环境的运行时检测框架，利用用户可访问的原始对数概率（logits）实现遗漏实时识别。实验基于5个模型、2个引擎共75,476次测试表明，总体遗漏率为0.62%，其中68%源于确定性中间件层（L0-L3），提示运维干预应聚焦于系统级配置而非模型本身。
链接: https://arxiv.org/abs/2607.22448
作者: Santhiya Rajan
机构: 未知
类目: Multiagent Systems (cs.MA)
备注:
Abstract:Air-gapped and on-premises deployments in regulated settings (clinical FHIR services, legal review, sovereign infrastructure) cannot call frontier APIs; they run quantized 4-8B models via this http URL or vLLM behind tool servers. The dominant reliability failure is omission: the silent absence of a decision-critical fact, such as an agent reading 20 of 400 records and reporting “no anomalies.” We argue omission is a pipeline phenomenon, not a model phenomenon, and make four contributions. First, a nine-layer taxonomy (L0-L8) locating every omission mechanism from ingestion through the agent loop. Second, an attribution methodology separating deterministic layers (L0-L3) from behavioral layers (L4-L8) via controlled ablation and logit decomposition, quantifying each with an omission waterfall. Third, an open cross-architecture harness comparing sliding-window-hybrid, full-attention, and SSM-hybrid models across engines and frameworks. Fourth, a runtime-detection framework for air-gapped settings where you own the logits. Results from a 75,476-trial sweep across five models and two engines show a pooled omission rate of 0.62; 68% originates in deterministic middleware (L0-L3), relocating where operators should intervene. Server-side profile factors (weight quantization, KV-cache type, RoPE scaling) were fixed and left for future work.
[MA-2] Draining the Energy Commons: Self-Defeating Over-Appropriation as a Coordination Failure in Agent ic LLM Collectives
【速读】：该论文旨在解决大语言模型（LLM）作为代理在共享公共资源（如可再生能源储备）时所面临的协调失败问题，特别是在电力生产与消费（prosumer）的自洽对弈环境中。当多个同源的LLM代理（如GPT、Gemini或Grok系列）在持续时间中共同使用有限的能源储备时，个体追求自身运营连续性的行为会引发集体资源过度消耗，导致未来服务能力下降。其解决方案的关键在于揭示：尽管各代理群体在需求未超过可再生能源最大替代水平时能自我约束保护资源，但一旦需求超出该阈值，便普遍出现资源过采现象（所有九组稀缺性对比均通过霍姆校正，最大调整后p值为4.87×10⁻⁵），形成“自我挫败”的动态——当前保护行为实质上破坏了未来的可持续性。进一步分析表明，在高稀缺性条件下（ρ=1.2），早期总请求压力普遍超过峰值可再生替代能力，平均达其1.21倍，且资源均值轨迹在第5至第7轮即低于最大再生水平。与社会最优规划者（最大化整体服务价值）和开放接入基准（个体私利最大化）相比，真实系统表现更接近于更具短期偏好的开放接入情境，表明这些代理在系统层面表现出类似“短视优化者”的行为特征。这一系统级对齐失效无法通过孤立响应评估发现，凸显了在共享资源场景下对多智能体长期动态进行整体评估的重要性。
链接: https://arxiv.org/abs/2607.22188
作者: Marcantonio Bracale Syrnicov,Federico Pierucci,Matteo Prandi,Marcello Galisai,Piercosma Bisconti,Francesco Giarrusso,Daniele Nardi
机构: 未知
类目: Multiagent Systems (cs.MA)
备注:
Abstract:LLMs are increasingly deployed as agents that plan, use tools, and act over time. When they share persistent resources, such as compute pools or energy reserves, decisions by one agent affect the conditions faced by later agents. We study this coordination failure in a renewable energy commons. Four same-family GPT, Gemini, or Grok agents act in homogeneous self-play as electricity prosumers, instructed to maximize operational continuity. Holding aggregate residual demand and the decision protocol fixed, we vary the regeneration rate of a shared energy reserve from abundance to scarcity. All three families preserve the reserve when demand does not exceed peak renewable replacement, but over-appropriate it beyond that threshold (all nine exact scarcity contrasts survive Holm correction; largest adjusted p = 4.87e-5). The pattern is self-defeating: the same populations protect current service while undermining future service. At higher scarcity (rho = 1.2), early aggregate request pressure exceeds peak renewable replacement in every family and averages 1.21 times that level. Mean trajectories fall below the reserve level of maximum replenishment by rounds 5-7. Two offline benchmarks compare a social planner maximizing group-wide operational-service value with open access, where each prosumer maximizes its own value. At a discount factor of gamma = 0.95, both benchmarks sustain the reserve under the same dynamics. Realized depletion instead resembles outcomes under a more impatient open-access benchmark. The populations therefore behave like impatient optimizers at the level of the public trajectory. This system-level alignment failure would be missed by isolated-response evaluation.
[MA-3] Predictive Lightweight MARL for Resilient Coverag e in Sparse-Signaling Aerial Networks
【速读】：该论文旨在解决带宽受限的无人机集群（UAV swarms）在面临稀疏信号传输与信息老化时，易出现协同失效（coordination collapse）的问题，从而保障覆盖的韧性。其核心解决方案是提出一种预测式轻量级多智能体强化学习（Predictive Lightweight Multi-Agent Reinforcement Learning, PL-MARL）框架，关键在于引入运动学感知推理引擎（Kinematic-Aware Inference Engine），通过利用物理先验知识主动重构邻近智能体的轨迹，实现对通信需求的显著降低。该方法在不依赖高频信号交换的前提下，实现了结构韧性与通信频率的解耦，有效平衡了计算开销与通信开销，大幅减少控制信令负载，从而为载荷业务保留更多频谱资源。仿真结果表明，该方案在极端信号稀缺及节点失效条件下仍能维持优异的覆盖性能与任务连续性，验证了主动推理机制在实现低延迟、可扩展的空中协同中的有效性。
链接: https://arxiv.org/abs/2607.22109
作者: Chuan-Chi Lai,Ang-Hsun Tsai
机构: National Chung Cheng University (国立中正大学); Advanced Institute of Manufacturing with High-tech Innovations (AIM-HI) (先进制造高技术研究中心); Feng Chia University (逢甲大学)
类目: Networking and Internet Architecture (cs.NI); Multiagent Systems (cs.MA); Systems and Control (eess.SY)
备注: Accepted for publication in IEEE Wireless Communications Letters. ©2026 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses
Abstract:This letter proposes the Predictive Lightweight Multi-Agent Reinforcement Learning (PL-MARL) framework to ensure resilient coverage in bandwidth-constrained UAV swarms. To counter coordination collapse caused by sparse signaling and information aging, we introduce a Kinematic-Aware Inference Engine that proactively reconstructs neighbor trajectories via physical priors. This approach enables an efficient computation-for-communication trade-off, decoupling structural resilience from signaling frequency. Simulations confirm that PL-MARL maintains superior coverage and mission continuity under extreme signaling scarcity and node failure. Our results validate proactive inference as a scalable, low-latency solution for robust aerial coordination, effectively minimizing control overhead to preserve spectrum for payload services while ensuring resilience against interference.
[MA-4] When Language Models Meet NeuroGraphs: Exploring Enhanced Agent ic LLM Framework Towards Brain Network Analysis
【速读】：该论文旨在解决现有深度学习方法在脑网络分析中将连接组分类简化为“图到逻辑值”的分类任务，导致解释性不足的问题。其核心挑战在于如何实现具备可解释性、知识增强且结构感知的脑网络分析。解决方案的关键是提出一种名为BrainAgent的智能体式大语言模型（LLM）框架，通过将连接组分类重构为一个迭代过程：包括拓扑感知理解、外部知识检索、推理与反思验证。该框架首先利用脑科学专用分析工具将原始脑网络转化为多层次结构描述，随后基于神经科学知识和特定任务案例进行知识锚定，最终生成带有反思验证的结构化预测结果。实验表明，BrainAgent在四个公开的静息态功能磁共振成像（rs-fMRI）数据集上均显著优于直接提示和标准推理基线，且消融与可解释性分析验证了各模块的有效性，证明了智能体式大语言模型在实现可解释、知识驱动的脑网络分析方面具有实际可行性。
链接: https://arxiv.org/abs/2607.22082
作者: Jiaxing Li,Rui Dong,Muyao Tang,Youyong Kong
机构: Southeast University (东南大学)
类目: Multiagent Systems (cs.MA)
备注:
Abstract:Brain network analysis is crucial for understanding cognition and neurological disorders, yet existing deep learning methods mainly treat connectome analysis as a graph-to-logit classification problem, offering limited explanatory reasoning. Large language models (LLMs) provide a promising interface for knowledge-intensive scientific analysis, but directly applying general-purpose LLMs to brain networks remains challenging due to the structure-language gap, limited neuroscience grounding, and overconfident positive predictions. In this paper, we propose \textbfBrainAgent, an agentic LLM framework for knowledge-enhanced brain network analysis. BrainAgent reformulates connectome classification as an iterative process of topology-aware understanding, external retrieval, reasoning, and reflection. Specifically, it first converts raw brain networks into compact multi-level structural descriptions through brain-specific analysis tools, then retrieves relevant neuroscience knowledge and task-specific cases to ground the reasoning process, and finally generates structured predictions with reflective verification. Experiments on four public rs-fMRI datasets show that BrainAgent consistently improves different closed-source and open-source LLM backbones over direct prompting and standard reasoning baselines. Further ablation and interpretability analyses demonstrate the effectiveness of each component and show that BrainAgent produces more comprehensive, multi-level, and verifiable this http URL results indicate that agentic LLMs provide a practical route toward interpretable and knowledge-grounded brain network analysis.
[MA-5] Reliability-Contagion Feasibility in LLM Multi-Agent Networks
【速读】：该论文旨在解决大语言模型（Large Language Model, LLM）代理在通过通信协作共享证据时，错误信息在异构网络中传播的风险问题。其核心挑战在于，在提升模型协同能力的同时，如何平衡通信带来的可靠性与错误扩散之间的矛盾。解决方案的关键是提出一种修正感知的网络传播模型（correction-aware network model），该模型借鉴流行病学中的四类动态（易感、暴露、感染、已修正），刻画错误信息在多智能体系统中的传播路径，并推导出在异构通信网络中错误早期入侵的临界条件。进一步地，该模型与基于多数投票的可靠性基准耦合，揭示了在固定每条通信边暴露程度下，可靠性目标与误差控制对网络连通性施加的相互冲突的约束。研究分析了这两种约束交集的存在性，识别出在特定条件下存在一个中间连通性范围可同时满足可靠性和抗错要求，并发现规则图在满足可靠性前提下能实现最小的入侵因子。此外，在固定发送方预算下，一阶阈值与网络密度无关，表明通信预算机制决定了新增边是否加剧早期传播风险。仿真与可控实验验证了上述理论预测：随着节点度从2增至5，平均首代后代数从0.667上升至1.667，而暴露邻居采纳率保持稳定，全级联场景下的非种子错误采纳率分别为0.200、0.333和0.333，证明了所提框架在连接性选择上的可解释性与实用性。
链接: https://arxiv.org/abs/2607.21912
作者: Ruiwu Niu,Xincheng Shu,Ying Zhao
机构: Hong Kong Shue Yan University(香港树仁大学); Beijing Normal University(北京师范大学); City University of Hong Kong(香港城市大学)
类目: Multiagent Systems (cs.MA)
备注:
Abstract:Communication allows large language model agents to pool evidence, but it also creates paths along which an erroneous claim can spread. We formulate a correction-aware network model that tracks susceptible, exposed, infectious, and corrected agents and derive its early-invasion condition for heterogeneous communication networks. We then couple this propagation model to an analytic majority-vote benchmark in which a clean-task reliability target imposes a minimum connectivity requirement. Under fixed exposure per communication edge, reliability and error control impose opposing graph constraints. We characterize when their intersection is empty and when it contains an intermediate connectivity range, and identify regular graphs that attain the smallest invasion factor within the reliable graph class when such graphs exist. Under a fixed sender budget, the homogeneous first-order threshold is independent of network density, showing that the communication-budget convention determines whether added edges increase early propagation risk. Finite-network simulations on 21,000 trajectories illustrate these directional predictions. A controlled grok-4.3 experiment then evaluates three six-node topologies on 36 new closed-world tasks, with a balanced 12-task subset continued to full cascades. Mean first-generation offspring increased from 0.667 to 1.333 and 1.667 as degree increased from 2 to 4 and 5, while the adoption fraction among exposed neighbours remained 0.333. Mean non-seed erroneous adoption in the full-cascade subset was 0.200, 0.333, and 0.333. Together, these results provide a tractable basis for selecting connectivity under explicit reliability and propagation constraints.
[MA-6] Equilibrium Causal Digital Twins: Validation Transport and Identification Limits
【速读】：该论文旨在解决数字孪生（Digital Twin）在具有反馈机制的系统中进行干预预测时的可验证性与跨域迁移性问题。核心挑战在于：当系统机制发生变化时，原本在某一领域构建的数字孪生模型可能无法准确预测目标场景下的反事实结果（counterfactual），尤其在涉及均衡选择的因果博弈中，传统方法难以保证预测的有效性。其解决方案的关键在于提出一套基于结构化假设的验证与迁移准则，具体包括：引入循环选择图（cyclic selection diagrams）以刻画机制依赖关系；建立直接复用与混合模型（hybrid models）的适用条件，其中结合源域的不变机制与目标域的信息；并通过一个不可能性结果证明，仅在有限实验设计下均符合观测分布的模型仍可能在目标反事实上产生分歧，从而强调了结构性假设的必要性。对于线性模型，进一步推导出依赖于变化机制、观测模型及图支持结构的干预设计要求，并在点识别失败时刻画查询值的可接受范围。此外，论文还提供了重构均值与协方差的统计检验方法，并通过合成反馈系统验证理论的有效性。
链接: https://arxiv.org/abs/2607.21667
作者: Faraz Dadgostari,Neda Nazemi
机构: Montana State University (蒙大拿州立大学); Gianforte School of Computing (Gianforte 计算机学院)
类目: Methodology (stat.ME); Multiagent Systems (cs.MA); Systems and Control (eess.SY); Optimization and Control (math.OC)
备注: Working Paper
Abstract:Digital twins are often used to predict how a system would respond to an intervention. In systems with feedback, a twin must reproduce an equilibrium counterfactual, and a twin developed in one domain may fail after mechanisms change. We study when these predictions can be validated and transported. For equilibrium causal games, we give conditions on the mechanisms, equilibrium selection, and intervention design under which agreement with experimental distributions identifies the counterfactual of interest. We show why agreement of means and covariances is insufficient for distributional queries. We then introduce cyclic selection diagrams and derive criteria for direct reuse and for hybrid models that combine invariant source mechanisms with target information. An impossibility result constructs systems that agree under every experiment in a finite design but disagree on the target counterfactual, showing that validation requires structural assumptions. For linear models, we derive intervention requirements that depend on the mechanisms that changed, the observation model, and graph support. When point identification fails, we characterize the remaining range of query values. We also provide statistical tests for reconstructed means and covariances and illustrate the theory in synthetic feedback systems.
自然语言处理
[NLP-0] Skill Self-Play: Pushing the Frontier of LLM Capability with Co-Evolving Skills
【速读】： 该论文旨在解决大语言模型（LLM）在自进化训练过程中面临的根本性矛盾：任务多样性与验证可靠性之间的权衡。现有方法中，环境受限的方法虽能获得精确反馈，但学习范围局限于狭窄领域；而开放式的自生成方法虽拓展了任务空间，却因缺乏可靠验证机制，易导致误导性奖励污染训练循环。为此，论文提出将“代理技能（agent skills）”作为破解该困境的关键中间路径：每项技能确保在特定场景下实现深度且可验证的执行，同时通过动态技能调度维持开放式的任务多样性。基于此洞察，作者提出了协同进化框架——技能自对弈（Skill Self-Play, Skill-SP），其由提议者（proposer）、求解者（solver）和动态技能控制器（dynamic skill controller）构成，三者通过强化学习循环协同演化：提议者基于动态采样的技能生成具有挑战性的任务；求解者探索候选解以突破自身能力边界；技能控制器则收集执行反馈以更新并扩展技能库。这一交互式协同进化机制有效弥合了结构化验证与开放式探索之间的鸿沟。实证结果表明，Skill-SP作为稳健的进化引擎，不仅能持续提升成熟模型的性能上限，还能显著逆转初始不匹配模型的表现，展现出强大的适应性与进化潜力。
链接: https://arxiv.org/abs/2607.22529
作者: Siyuan Huang,Pengyu Cheng,Haotian Liu,Tao Chen,Yihao Liu,Jingwei Ni,Shijie Zhou,Ziyi Yang,Gangwei Jiang,Mengyu Zhou,Yu Cheng,Xiaoxi Jiang,Guanjun Jiang
机构: 未知
类目: Computation and Language (cs.CL)
备注:
Abstract:LLM training is shifting from manual design and annotation to interaction-driven self-evolution. However, existing self-evolutionary methods face a fundamental dilemma between task diversity and verification reliability: environment-bound methods obtain precise feedback but confine learning to narrow domains, while open-ended self-generation broadens the task space but lacks reliable verification, allowing misleading rewards to pollute the training loop. We identify agent skills as a powerful middle ground to reconcile this tension: each skill ensures deep, verifiable execution in a specific scenario, while dynamic routing across skills maintains open-ended task variety. Leveraging this insight, we introduce Skill Self-Play (Skill-SP), a co-evolutionary framework comprising a proposer, a solver, and a dynamic skill controller. Orchestrated via a reinforcement learning loop, these components co-evolve in a continuous self-play loop: the proposer generates challenging tasks conditioned on dynamically sampled skills; the solver explores candidate solutions to push its capability boundaries; and the skill controller collects execution feedback to update and expand the skill library. This interactive co-evolution effectively bridges the gap between structured verification and open-ended exploration. Empirical evaluations on tool-use and reasoning benchmarks demonstrate that Skill-SP, serving as a robust evolution engine, consistently pushes the performance ceiling of competent backbones while catalyzing striking turnarounds for initially misaligned models. Our code is available at this https URL.
[NLP-1] Opaque Epistemic Mediation: How LLM Deployment Configurations Shape the Validation of Pseudo-Science
【速读】： 该论文旨在解决商业大语言模型（Large Language Models, LLMs）在面对有争议的科学主张时，其认知立场缺乏稳定性与透明性的问题。具体而言，研究聚焦于评估主流LLM家族（Claude、Grok、GPT、Gemini）对基于弗兰克·萨尔特（Frank Salter）生物社会框架的种族民族伪科学（ethnonationalist pseudo-science）的可信度判断，并考察其随时间与部署环境变化的动态表现。研究发现，Grok的快速版本在默认X平台体验中持续赋予此类伪科学高达70–75的可信度评分，显著高于其他模型（15–40），而这一异常行为在对照任务（如演化共识或已被证伪的拉马克主义）中并未出现，表明其响应并非源于模型本身的知识结构，而是特定部署配置所致。关键发现包括：（1）一次无声补丁使Grok的行为从混乱变为稳定高可信度，且无公开记录；（2）同一模型标识在不同接口（API vs. Web）下输出差异巨大（75 vs. 5.5）；（3）最合理的拒评行为（即拒绝评估伪科学）在部分模型及其后续版本中逐渐消失。这些结果表明，商业大语言模型的认知立场并非模型固有的稳定属性，而是由系统提示词、安全层、接口路由及无声更新等部署配置所决定的偶然性结果，且对用户和研究者均不透明。因此，研究强调必须建立新型的“认识论问责”机制，以应对生成式人工智能在知识权威性判断中的不可靠性与隐蔽性风险。
链接: https://arxiv.org/abs/2607.22513
作者: Davide Scarso,Hugo Noronha de Almeida,Joaquim Pina
机构: 未知
类目: Computers and Society (cs.CY); Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注: 16 pages, 2 tables
Abstract:Commercial large language models are increasingly used as knowledge references, yet their stance on contested scientific claims is neither stable nor transparent. We tested how four major LLM families (Claude, Grok, GPT, Gemini) evaluate ethnonationalist pseudo-science derived from Frank Salter’s biosocial framework across four temporal snapshots (October 2025-February 2026), via both API and web interfaces. Grok’s Fast versions (which power the default user experience on X) consistently assigned credibility scores of 70-75, two to five times higher than all other models (which scored 15-40). This pattern was absent from control prompts testing basic evolutionary consensus and refuted Lamarckian claims, where all models performed comparably. Three additional findings emerged: (1) a silent patch reversed Grok’s behaviour from chaotic to stably high validation overnight, without any public documentation; (2) the same Grok model identifier produced radically divergent outputs via API (75) and web (5.5) three months later; (3) refusal to rate the pseudo-scientific claim, the most defensible response observed, appeared in two model families through different interfaces (Claude Opus 4.1 categorically via web, GPT-5.1 Chat intermittently via API) and eroded in the successor version of each. These results indicate that the epistemic stance of a commercial LLM is not a stable property of the model but a contingent effect of deployment configuration: system prompts, safety layers, interface routing, and silent updates. This remains opaque to users and researchers alike. We argue this constitutes a matter of public concern requiring new forms of epistemic accountability.
[NLP-2] grapheme-kit: Grapheme-Level Metrics and Text Processing for Multilingual NLP
【速读】： 该论文旨在解决现有词汇距离、相似性及评估度量方法在处理复杂书写系统时的局限性问题，即这些方法基于Unicode码点进行计算，而未能准确反映实际书写中单个音位（grapheme）可能由多个Unicode码点组成的事实，从而导致对文本错误的误判。其解决方案的关键在于提出并实现grapheme-kit这一开源Python库，该库将传统度量扩展至以音位簇（grapheme cluster）为基本单位，实现了对音位簇的精确识别与处理，并特别优化了泰米尔语和僧伽罗语的音位处理能力，包括音位的组合与分解功能。通过光学字符识别（OCR）案例研究验证，基于音位层级的度量能够更真实地评估复杂脚本的性能。
链接: https://arxiv.org/abs/2607.22456
作者: Izzath Nisfer,Ashini Kavindya,Ovindu Atukorala,Purushoth Velayuthan,Menan Velayuthan
机构: 未知
类目: Computation and Language (cs.CL)
备注:
Abstract:Existing lexical distance, similarity, and evaluation metrics operate on Unicode code points, which can misrepresent errors in writing systems where a single grapheme is represented by multiple Unicode code points. We introduce grapheme-kit, an open-source Python library that extends these metrics to operate on grapheme clusters instead. The library also provides improved grapheme processing for Tamil and Sinhala, including accurate grapheme cluster identification and grapheme composition/decomposition utilities. Through an OCR case study, we demonstrate that grapheme-level metrics provide a more faithful evaluation of complex scripts.
[NLP-3] A Factorial Study of Synthetic Data Generation for Low-Resource Machine Translation using Grammar Books
【速读】： 该论文旨在解决濒危语言因缺乏平行语料而无法实现机器翻译的问题，尽管这些语言已有描述性语法书等语言学资源。其核心解决方案是构建一个端到端的流水线，利用大语言模型（Large Language Models, LLMs）从语法书中自动提取句法规则、例句和词典，并生成合成的平行语料用于模型微调。与以往在推理时将语法内容作为提示（prompt）输入的方法不同，本方法将语法知识转化为可训练的合成数据，从而在训练阶段充分整合语言学先验知识。在三种语言类型差异显著的低资源语言（卡拉芒语，Papuan；图茨钦语，Romance；曼丹语，Siouan）上的实验表明，基于合成数据的微调在75%的配置中优于种子数据基线，最高分别取得+8.8、+5.3和+3.3的ChrF++提升。通过系统性的96种配置因子组合（目标词性、检索粒度、样本量）分析，研究识别出影响性能的关键因素组合及其失效边界。结果表明，静态的语言学文献可通过自动化流程转化为可用于机器翻译微调的有效数据，为严重资源匮乏语言提供了可行的翻译工具开发路径。
链接: https://arxiv.org/abs/2607.22376
作者: Varun Ghat Ravikumar,Sina Ahmadi,Lena Jäger,Rico Sennrich
机构: University of Zurich (苏黎世大学)
类目: Computation and Language (cs.CL)
备注: Accepted at CLiC-it 2026
Abstract:Most endangered languages lack the parallel data required for machine translation, despite the existence of descriptive grammar books. We introduce a pipeline that uses large language models to extract grammatical rules, example sentences, and lexicons from grammar books and generate synthetic parallel corpora for fine-tuning-rather than feeding grammar content into prompts at inference time, as in prior work. Validated on three typologically diverse low-resource languages-Kalamang (Papuan), Tuatschin (Romance), and Mandan (Siouan)-we show that fine-tuning on synthetic data improves over seed-data baselines in 75% of configurations for Kalamang and 59% for Tuatschin, with best-case ChrF++ gains of +8.8, +5.3, and +3.3 respectively. Through a systematic factorial study across 96 configurations varying target part-of-speech, retrieval granularity, and sample volume, we identify which factor combinations drive gains and where they break down. Our results demonstrate that static linguistic documentation can be repurposed for machine translation fine-tuning, offering a practical path towards translation tools for severely under-resourced languages.
[NLP-4] Cross-Tokenizer On-Policy Distillation via Byte-Prefix Marginalization
【速读】： 该论文旨在解决多源异构语言模型（如不同家族的开源大模型）在进行知识蒸馏时，因分词器不一致导致的教师-学生词汇表对齐难题。传统跨分词器方法常面临概率质量丢失或错误分配至语义无关的学生词元的问题，从而影响蒸馏效果。其解决方案的关键在于提出一种名为字节前缀归一化（Byte-Prefix Marginalization, BPM）的新方法，通过在共享字节空间中重构教师模型的下一个词元分布，实现对齐与质量保留。BPM将每个教师词元的概率分配给其字节表示为前缀的最长学生词元，聚合相同学生词元所接收的概率质量，并将未匹配部分显式归入残差类别，从而构建一个完整覆盖、字节对齐且质量守恒的目标分布。该方法在超过99%的训练位置上能精确恢复教师诱导的字节前缀边际分布，在其余情况下采用保质的链式因子化下界近似。实验表明，基于Qwen3-32B、GLM-Z1-9B-0414和MiniMax-M2.7等教师模型，BPM在六项数学与编程基准测试中显著优于现有方法，平均性能提升达3.7–6.6点。
链接: https://arxiv.org/abs/2607.22334
作者: Hao Wang,Kun Yuan,Wenlin Zhong,Minglei Zhang,Han Xiao,Ming Sun,Honggang Qi
机构: 未知
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注: Project page: this https URL
Abstract:Open-weight language models from different families exhibit complementary capabilities, motivating their consolidation into a compact student through on-policy distillation (OPD). However, full-vocabulary OPD typically assumes a shared tokenizer, while existing cross-tokenizer methods may discard teacher probability mass or assign it to student tokens with unrelated content. We introduce Byte-Prefix Marginalization (BPM), which re-expresses the teacher’s next-token distribution over the student vocabulary in a shared byte space. Specifically, BPM assigns each teacher token’s probability to the longest student token whose byte representation is a prefix of the teacher token’s bytes, aggregates mass mapped to the same student token, and places otherwise unmatched mass in an explicit residual category. This produces a vocabulary-complete, byte-aligned, and mass-preserving target for dense OPD. The target exactly recovers the teacher-induced byte-prefix marginal when the relevant prefix does not span multiple teacher tokens (a condition satisfied at more than 99% of training positions) and uses a mass-preserving, chain-factorized lower bound otherwise. Across Qwen3-32B, GLM-Z1-9B-0414, and MiniMax-M2.7 as teachers, BPM consistently outperforms current cross-tokenizer methods on six mathematics and programming benchmarks, improving six-benchmark avg@8 by 3.7-6.6 points over the strongest baselines.
[NLP-5] Biomedical Machine Translation for Low-Resource Arabic-Script Languages via Cross-Lingual Transfer and LoRA Adapter Merging
【速读】： 该论文旨在解决阿拉伯语系低资源语言在生物医学领域机器翻译（Biomedical NMT）资源极度匮乏的问题，尤其聚焦于达里语（Dari）、普什图语（Pashto）、苏拉尼库尔德语（Sorani Kurdish）和乌尔都语（Urdu）等严重低资源目标语言的跨语言迁移难题。其核心解决方案是利用阿拉伯语和波斯语作为高资源中间语言（pivot languages），通过在小型解码器仅模型（decoder-only LLMs）上采用低秩适应（LoRA）微调技术，训练特定领域的中间适配器（domain-specific pivot adapters），并系统评估三种迁移策略：少样本上下文学习、最小监督微调以及首次应用于该场景的零数据LoRA适配器合并（zero-data LoRA adapter merging）。关键创新在于，仅需500句标注数据的监督微调即可使达里语翻译质量接近中间语言水平（CHrF++ 41.01），对乌尔都语也取得显著提升（CHrF++ 28.88）；而零数据适配器合并方法在无需额外目标语言数据的情况下，使达里语性能逼近监督微调结果（差距仅3.5 CHrF++），展现出极高的效率与潜力。然而，当目标语言与中间语言结构差异过大时（如普什图语和苏拉尼库尔德语），现有跨语言迁移方法仍难以满足临床部署的高要求，揭示了当前方法在处理远距离语言迁移时的局限性。
链接: https://arxiv.org/abs/2607.22300
作者: Abdullah Alabdullah,Arash Eslamighayour,Sarp Harbalioglu,Lifeng Han
机构: 未知
类目: Computation and Language (cs.CL)
备注:
Abstract:We present a systematic study of healthcare-domain cross-lingual transfer to address the scarcity of biomedical NMT resources for Arabic-script languages. We use Arabic and Persian as higher-resource pivots to improve translation for \textbffour severely low-resource targets: Dari (Afghan Persian, a standardised variety of Persian), Pashto, Sorani Kurdish (Central Kurdish, a major standardized variety of Kurdish), and Urdu (closely related to Hindi). Using LoRA fine-tuning on small decoder-only LLMs, we train \textitdomain-specific pivot adapters and evaluate \textbfthree transfer strategies: few-shot in-context learning, minimal supervised adaptation, and, to the best of our knowledge, for the first time in this setting, zero-data LoRA adapter merging. Supervised adaptation with just 500 sentences achieves near pivot-language quality for Dari (CHrF++ 41.01) and meaningful gains for Urdu (28.88), while adapter merging reaches within 3.5 CHrF++ of supervised adaptation for Dari at zero additional cost. Pashto and Sorani Kurdish remain insufficient for high-stakes clinical deployment exposing the limits of cross-lingual transfer when structural distance from the pivots is too great. LoRA adapter merging works surprisingly well for closely related languages, even without target-language biomedical data.
[NLP-6] Why Large Language Models and Humans Converge and Diverge in Evaluating Creativity
【速读】： 该论文旨在解决生成式 AI（Generative AI）在创意评估中与人类判断的对齐问题，具体探究在何种条件下大语言模型（Large Language Models, LLMs）的创意评价会与人类一致或产生偏差。其核心问题是：当前广泛应用的LLMs在评估创意时所依据的标准是否与人类一致？解决方案的关键在于系统识别出不同LLMs用于创意评价的内在标准，并揭示这些标准差异如何影响实际评价结果。研究发现，尽管所有LLMs均在新颖性（novelty）维度上与人类判断有较强一致性，但在涉及社会、市场及声誉等上下文信息的维度上存在显著分歧；此外，各模型表现出独特且差异显著的评价标准范围。这种标准差异直接反映在评价结果中：拥有更广泛标准的模型能更好地区分人类认为更具创造力的想法，而多数模型对上下文信息不敏感——即便这些信息显著改变人类评分，其评分仍基本不变。因此，研究揭示了LLM与人类在创意评价上的对齐程度取决于评价任务所依赖的信息类型及其模型自身采用的标准体系，强调选择特定评估模型本质上是基于其内在标准对“创造性”进行定义的决策过程。
链接: https://arxiv.org/abs/2607.22218
作者: Pengzhao Lyu,Yeun Joon Kim,Hanlin Xiao,Yingyue Luna Luan
机构: 未知
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注:
Abstract:Despite the growing use of large language models (LLMs) as creativity evaluators, evidence of their alignment with human evaluations remains mixed, raising the question of when and why their judgments converge with or diverge from human judgments. Across three studies and six widely used LLMs, we addressed this gap by identifying the standards underlying LLM creativity evaluation and examining their downstream implications. Study 1 showed that LLMs generally relied on a narrower subset of human creativity evaluation standards. Convergence with human standards was strongest in the novelty dimension, whereas divergence was clearest in the contextual dimension, which captures social, market, and reputational information. Moreover, each LLM exhibited distinct, model-specific standards that varied substantially in breadth. These differences in evaluation standards were reflected in actual creativity judgments. Study 2 (N = 1,103 ideas) showed that LLM evaluations were moderately correlated with human evaluations, and individual LLMs with broader standards better distinguished ideas humans judged as more versus less creative. Study 3 (N = 1,195) showed that LLMs were less sensitive to contextual information: such information significantly altered human creativity ratings but left LLM ratings largely unchanged. Together, our findings help explain the mixed evidence on LLM-human alignment, showing that alignment depends on the evidence a judgment demands and the standards each model applies. LLMs may resemble humans when evaluations emphasize intrinsic qualities such as novelty, yet diverge when judgments require contextual information. Selecting an LLM evaluator is therefore a consequential decision: different models, applying different standards, recognize different ideas as creative.
[NLP-7] From Isolated Tasks to Structured Capabilities: A Multilayer Taxonomy for Large Language Models
【速读】： 该论文旨在解决大语言模型（Large Language Model, LLM）评估中存在任务碎片化的问题，即当前评估体系以具体任务为导向，缺乏对所探查能力的系统性组织，导致跨研究比较困难、能力覆盖不全难以识别。其核心解决方案是构建一个包含14个能力域与91个子技能的多层分类体系，该体系分为原始（Primitive）、建构（Constructed）和整合（Integrative）三层，基于人类认知科学理论定义能力结构，而非依附于模型架构。层级划分依据发展优先顺序及功能支持假设，并将人类认知概念适配为可观测的模型行为。通过在2023–2025年间对ACL、AAAI、ICML和NeurIPS共31,505篇论文进行多模型标注、共识与仲裁，实现了15,934篇聚焦LLM的研究文献映射。结果显示，研究注意力高度集中于语言-语义能力（22.3%）、推理（21.3%）、规划与决策（13.5%）和感知（12.3%），而部分能力域（如心智理论与社会推理互动）虽出现频率低，但表现出极高共现提升值（lift=30.84），揭示潜在重要性。该分类体系将分析单位从孤立任务转向结构化能力，显著提升了研究组织、覆盖审计、评估解释以及可验证诊断、训练与迁移假设的能力。
链接: https://arxiv.org/abs/2607.22182
作者: Shixin Fang,Jiachen Wo,Wenjuan Qin,Sihang Jiang,Yanghua Xiao(Fudan University)
机构: Fudan University (复旦大学)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: 34 pages, 5 figures, 20 tables
Abstract:Large language model (LLM) evaluation spans diverse tasks and benchmarks, yet evidence remains organized around tasks rather than the capabilities they probe. This fragmentation limits cross-study comparison, obscures capabilities tasks recruit, and makes coverage gaps difficult to identify. We introduce a multi-layer taxonomy of 14 capability domains and 91 subskills across Primitive, Constructed, and Integrative layers. Human cognitive science guides capability definition and organization, not LLM architecture. Layer assignments draw on developmental precedence and hypothesized functional support, while human-origin constructs are adapted to observable model behavior. To demonstrate operational utility, we screened 31,505 papers from ACL, AAAI, ICML, and NeurIPS between 2023 and 2025 and mapped 15,934 LLM-focused papers through multi-model annotation, consensus, and arbitration. Direct research attention concentrated on Language-Semantic Competence (3,551; 22.3%), Reasoning (3,388; 21.3%), Planning and Decision-Making (2,149; 13.5%), and Perception (1,954; 12.3%), whereas six domains appeared in fewer than 2% of papers. Within domains, the most frequent subskill had a median prevalence of 97.9% and appeared in at least 90% of papers in 10 of 14 domains. Language-Semantic Competence and Reasoning formed the highest-volume pair (n = 1,864; 11.7%; lift = 2.47), whereas Theory of Mind and Social Reasoning and Interaction showed the highest lift among pairs with at least 20 co-occurrences (n = 62; lift = 30.84). By shifting the unit of analysis from isolated tasks to structured capabilities, the taxonomy supports research organization, coverage audits, evaluation interpretation, and testable hypotheses for diagnosis, training, and transfer. Comments: 34 pages, 5 figures, 20 tables Subjects: Computation and Language (cs.CL); Artificial Intelligence (cs.AI) ACMclasses: I.2.7; I.2.6 Cite as: arXiv:2607.22182 [cs.CL] (or arXiv:2607.22182v1 [cs.CL] for this version) https://doi.org/10.48550/arXiv.2607.22182 Focus to learn more arXiv-issued DOI via DataCite (pending registration) Submission history From: Jiachen Wo [view email] [v1] Fri, 24 Jul 2026 10:49:50 UTC (1,374 KB)
[NLP-8] DBA-Bench: A Production-Fidelity Benchmark for LLM -Based Database Operations Agents
【速读】： 该论文旨在解决大语言模型（Large Language Model, LLM）驱动的数据库智能体在评估与实际生产运维之间存在的显著差距问题。现有评估体系受限于任务范围不一致、测试环境差异及评价指标多样化，难以真实反映智能体在复杂生产场景中的能力。核心问题包括：运行环境保真度不足（缺乏对活跃数据库的多轮读写交互模拟）、观测空间规模与复杂性缺失（无法覆盖数千条时间序列、业务日志及并发活动的因果诊断需求）、解决方案空间封闭（仅支持单一修复路径而忽略不同操作策略间的权衡）以及场景复杂性与覆盖度有限（难以模拟跨内部机制与运维域的故障级联）。为此，论文提出DBA-Bench基准，其关键在于通过高保真生产环境模拟、以结果为导向的评估范式以及可复现的受控场景配置来弥合上述差距。该基准采用注入监控的PostgreSQL实例，持续负载与持久状态保持，并融合多源观测数据；以满足安全约束下的可衡量恢复或故障消除作为成功标准；并通过在每轮测试前回滚至预设快照并执行场景特异性校验，确保实验一致性。该基准包含7个任务领域共106个场景，并提供基于参考路径诊断深度和环境复杂性的两个公开难度标签。在848次自动化测试中，诊断成功率、结果达成率与安全通过率分别为32.7%、19.6%和12.4%，最佳自动化基线仅达到17.9%的安全通过率，远低于人类数据库管理员（Human DBA）的93.4%表现，且在高难度场景下安全通过率下降至7.6%，凸显了实现安全、端到端自动修复的挑战性。
链接: https://arxiv.org/abs/2607.22165
作者: Junming Chen,Junyang Jiang,Xu Chen,Zibo Liang,Kai Zheng
机构: University of Electronic Science and Technology of China (电子科技大学)
类目: Databases (cs.DB); Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Machine Learning (cs.LG)
备注: 14 pages, 6 figures, 2 tables
Abstract:LLM-based database agents show promise, but differing task scopes, testbeds, and metrics hinder comparison. We identify four gaps between evaluation and production operations: live-environment fidelity (multi-turn read-write interaction with a running database); observation-space scale and complexity (causal diagnosis across thousands of time series, business logs, and concurrent activity); solution-space openness (multiple remediations with different operational trade-offs); and scenario complexity and coverage (faults cascading across internal mechanisms and operational domains). We present DBA-Bench, a benchmark addressing these gaps through production fidelity, outcome-first evaluation, and controlled scenario reproducibility. It uses instrumented PostgreSQL environments with active workloads, persistent state, and multi-source observations; defines success by measurable recovery or fault elimination under safety constraints; and restores snapshots with scenario-specific checks before each run. The benchmark contains 106 scenarios across seven task domains, with two public difficulty labels based on reference-path diagnostic depth and environmental complexity. We evaluate nine baseline groups, including six foundation-model systems, two GPT-5.5-backed database agents, and a Human DBA reference. Across 848 automated runs, Diagnosis, Outcome, and Safe Pass rates are 32.7%, 19.6%, and 12.4%; the best automated baseline reaches 17.9% Safe Pass versus 93.4% for the Human DBA reference. Automated Safe Pass falls from 19.6% on Easy scenarios to 7.6% on Hard scenarios, underscoring the difficulty of safe end-to-end remediation.
[NLP-9] Dynamic Commonsense Coordination for Empathetic Response Generation
【速读】： 该论文旨在解决情感化对话生成（Empathetic Response Generation, ERG）中因固定共知知识（commonsense knowledge）表示复用而导致的跨阶段知识协调能力不足的问题。现有方法在理解与生成阶段普遍采用静态共知知识表示，难以动态适配上下文语境，限制了情感识别与响应生成之间的协同效果。为此，论文提出动态共知知识协调框架（Dynamic Commonsense Coordination, DCC），其核心在于三个互补模块：基于残差的共知交互（SCE-AttnRes）用于融合上下文与情境相关的共知表征；关联引导的共知过滤（AGCF）通过抑制低相关性共知关系提升知识精准度；迭代式共知感知解码（ICAD）则在生成过程中动态检索共知记忆，实现知识的时序性调用。实验结果表明，DCC在Empathetic-Dialogues基准上显著提升了情绪分类准确率与响应多样性，同时保持与基线相当的困惑度，并在基于大语言模型（LLM）的盲评中展现出更优的相关性、连贯性与信息量。
链接: https://arxiv.org/abs/2607.22136
作者: Zhengyu Qi
机构: Leiden University (莱顿大学)
类目: Computation and Language (cs.CL)
备注:
Abstract:Empathetic Response Generation (ERG) requires models to recognize users’ emotions and generate empathetic responses. Commonsense knowledge has been shown to support such reasoning, yet existing approaches typically reuse fixed commonsense representations across understanding and generation, limiting their ability to coordinate such knowledge across different stages. We propose DCC, a Dynamic Commonsense Coordination Framework with three complementary modules: residual-based commonsense interaction (SCE-AttnRes) to integrate contextual and situational commonsense representations, Association-Guided Commonsense Filtering (AGCF) to down-weight low-relevance commonsense relations, and Iterative Commonsense-Aware Decoding (ICAD) to dynamically retrieve commonsense memories during generation. Experiments on the Empathetic-Dialogues benchmark show that DCC improves emotion classification accuracy and response diversity over the CEM baseline while maintaining comparable perplexity. An LLM-based blind evaluation further demonstrates that DCC generates responses with better relevance, coherence, and informativeness. The code and implementation details will be publicly available at this https URL.
[NLP-10] MEUSLI: a Multilingual Projector for LLM -based ASR and Beyond
【速读】： 该论文旨在解决现有轻量级语音投影器（Lightweight Projector）系统在多语言支持方面的局限性，尤其是其普遍仅支持少数语言（如英语）而难以覆盖低资源语言的问题。当前方法难以实现真正意义上的全开源端到端自动语音识别（ASR）在多语言场景下的部署，制约了开放科学与包容性语音理解技术的发展。本文提出的解决方案——MEUSLI，是首个基于Whisper语音编码器与开源多语言大语言模型（LLM）的开源多语言投影器家族，实现了28种欧洲语言的全开源端到端ASR。其关键技术在于通过持续学习（continual learning）策略，使投影器能够高效地扩展至训练中未见的新语言，同时在高资源与低资源语言上均取得优异性能。此外，研究进一步验证了MEUSLI在跨任务迁移中的泛化能力，仅需每语言数小时的任务特定微调，即可支持多语言语音翻译与话题识别等下游任务。因此，MEUSLI为构建可扩展、包容性强的开源语音大模型（SpeechLLM）提供了坚实基础。
链接: https://arxiv.org/abs/2607.22100
作者: Lorenzo Concina,Seraphina Fong,Marco Matassoni,Alessio Brutti
机构: Center for Augmented Intelligence, Fondazione Bruno Kessler(意大利布罗诺·凯斯勒基金会增强智能中心); Department of Information Engineering and Computer Science, University of Trento(特伦托大学信息工程与计算机科学系)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Audio and Speech Processing (eess.AS)
备注:
Abstract:Lightweight projectors are an established way to connect pre-trained speech encoders with large language models (LLMs), mapping acoustic features into token-level embeddings for tasks like ASR and spoken question answering. Existing systems, however, typically only support a few languages and are often limited to English. We introduce MEUSLI, the first open-science multilingual projector family that links a Whisper encoder with open-source multilingual LLMs, enabling fully open-source end-to-end ASR in 28 European languages. MEUSLI extends prior monolingual pipelines, delivering strong results across high- and low-resource languages. Using proper continual leaning techniques, MEUSLI can be easily extended to other languages not seen in training. We further demonstrate that the MEUSLI projector can be leveraged beyond ASR, enabling multilingual speech translation and topic identification with only a few hours of task specific supervision per language. Overall, MEUSLI provides a solid foundation for multilingual speech understanding tasks, supporting scalable and inclu- sive open-source SpeechLLM
[NLP-11] Nanbeige4.2-3B: Unlocking Agent ic Capabilities in a Compact Mode
【速读】： 该论文旨在解决小型化生成式智能体（generative agent）在保持高效推理能力的同时，如何在复杂工具使用、代码生成及办公自动化等任务中实现与更大模型相当甚至更优性能的问题。其核心解决方案在于：首先，通过从零开始预训练28万亿词的语料，并采用循环式变压器（Looped Transformer）架构，在不增加参数量的前提下提升模型容量；其次，构建多样化可执行环境、任务资产与智能体框架，结合真实部署与大规模合成数据增强监督微调（SFT）与轨迹生成的质量；再次，设计混合模式强化学习人类反馈（mixed-mode RLHF），对“思考”与“非思考”响应进行联合优化，以提升整体表现并减少失败案例；同时引入长度可控的推理强化学习，平衡推理准确性与效率；最后，采用基于结果与过程双重奖励的智能体强化学习策略，有效稳定长周期任务训练。实验证明，该模型在多个智能体基准测试中超越了如Qwen3.5-9B和Gemma4-12B等更大规模模型，且在本地化个人助理场景中表现出色，具备高实用性。
链接: https://arxiv.org/abs/2607.22083
作者: Nanbeige Lab:Chen Yang,Chengrui Huang,Fufeng Lan,Hanhui Chen,Hao Zhou,Huatong Song,Jiaqi Cao,Jiaying Zhu,Jinlin Niu,Kai Wang,Lisheng Huang,Qiliang Liang,Ran Le,Ruixiang Feng,Shuang Sun,Tao Gu,Tao Zhang,Tianyu Luo,Yang Song,Yun Xing,Yuntao Wen,Ziyao Xu,Zongchao Chen,Zongqiang Li
机构: Nanbeige LLM Lab(纳米贝语言模型实验室); Boss Zhipin( boss直聘)
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注:
Abstract:We present Nanbeige4.2-3B, a compact general agentic model with 3B non-embedding parameters. It delivers strong performance across code-agent, office-agent, and complex tool-use tasks while maintaining highly competitive reasoning capabilities in mathematics, coding, and science. Nanbeige4.2-3B is pretrained from scratch on 28T tokens with a Looped Transformer that reuses the layer stack to increase capacity without adding parameters. For SFT data and trajectory construction, we expand the diversity of executable environments, task assets, and agentic scaffolds through real-world deployment and large-scale synthesis. Our RL pipeline applies mixed-mode RLHF over Think and Non-Think responses to improve overall model quality and reduce failure cases, length-controlled reasoning RL to balance accuracy and reasoning efficiency, and agentic RL with outcome and process rewards to stabilize long-horizon training. Extensive evaluations show that Nanbeige4.2-3B outperforms larger models, including Qwen3.5-9B and Gemma4-12B, across diverse agentic benchmarks while remaining competitive on reasoning and alignment tasks. Performance with OpenClaw further supports its use as a compact local personal assistant.
[NLP-12] FSE: Continual Learning for Named Entity Recognition by Fast-Slow Experts
【速读】： 该论文旨在解决持续学习中的命名实体识别（Continual Learning for Named Entity Recognition, CLNER）问题，核心挑战在于模型在增量学习新实体类型时容易发生灾难性遗忘，且难以有效利用跨任务间的共享信息。其解决方案的关键是提出一种基于快速-慢速专家（Fast-Slow Experts, FSE）的端到端跨度感知神经网络架构：其中，共享的快速专家（Fast Expert）通过学习词级别关联，高效过滤掉低可能性的候选跨度，显著降低后续分类负担；而特定任务的慢速专家（Slow Expert）仅对剩余候选进行精细化的跨度分类，从而在保持模型可塑性的同时增强跨任务知识共享能力。此外，引入长度衰减负采样策略以缓解跨度数量不平衡问题。实验结果表明，FSE在OntoNotes和FewNERD合成数据集上均达到当前最优性能，验证了各组件的有效性、更快的收敛速度以及双专家机制的协同功能。
链接: https://arxiv.org/abs/2607.22075
作者: Yunan Zhang,Yang Fan,Heng Li,Xiangping Wu,Qingcai Chen
机构: Harbin Institute of Technology (哈尔滨工业大学)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: Preprint submitted to Pattern Recognition Letters
Abstract:Continual Learning for Named Entity Recognition (CLNER) enable models to incrementally learn new entity types without forgetting previously acquired ones. However, existing methods suffer from catastrophic forgetting and insufficient exploitation of shared information across tasks. This paper proposes FSE, a Fast-Slow Experts enhanced span-based NER model for CLNER. The shared fast expert learns token-level links to efficiently filter out unlikely spans, while the task-specific slow expert performs span classification only on the remaining candidates. It stabilizes learning by promoting knowledge sharing across tasks and maintains plasticity by reducing learning burden at each task. A length-decay negative sampling strategy to mitigate span imbalance is also introduced. Extensive experiments on OntoNotes and FewNERD synthestic datasets demonstrate that FSE achieves state-of-the-art performance in CLNER scenarios, with effectiveness of each component, empirical evidence of faster convergence and expected functionality of both experts.
[NLP-13] Benchmarking Fine-tuning and Retrieval Strategies for a Multimodal Language Model on the NRC Reactor Operator Licensing Examination
【速读】： 该论文旨在解决生成式人工智能（Generative AI）在核能领域应用中缺乏领域专业知识支撑的问题，即如何使大语言模型（LLM）具备符合核电厂操作员资质认证要求的专业知识应用能力。其核心解决方案在于通过结合监督微调（SFT）、检索增强生成（RAG）与检索增强微调（RAFT）等技术，并系统评估不同检索策略（固定大小滑动窗口与结构感知分块）对模型性能的影响。研究发现，采用基于Gemini蒸馏链式思维（CoT）推理的监督微调配合固定大小分块的RAG配置，在14份通用基础考试中成功达到80%人类通过率标准，表现最优；而未经过微调的配置均未能通过任何考试。此外，研究揭示了模型训练状态影响最佳分块策略选择的规律性现象，且在匹配搜索环境时，RAFT方法性能劣于标准SFT。结果表明，融合特定领域知识检索与针对性微调的组合策略是实现核电厂操作员级能力的关键。
链接: https://arxiv.org/abs/2607.22067
作者: Isak Hwang,Yoon Pyo Lee
机构: 未知
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注:
Abstract:The integration of large language models (LLMs) into the nuclear power industry requires outputs grounded in domain-specific knowledge. This study evaluates a 31-billion-parameter open-weight multimodal model (Gemma 4 31B-IT) on its capacity to apply nuclear knowledge by benchmarking eight model-retrieval configurations against the U.S. Nuclear Regulatory Commission (NRC) Reactor Operator licensing examination. We evaluate 14 Generic Fundamentals Examinations (GFE) from the 2015-2021 March sittings (seven pressurized and seven boiling water reactor exams) using the standard 80% human passing criterion. The base model is compared against configurations utilizing supervised fine-tuning (SFT) on Gemini-distilled chain-of-thought (CoT) rationales, retrieval-augmented generation (RAG) with BM25 sparse retrieval over the U.S. Department of Energy Fundamentals Handbook, and retrieval-augmented fine-tuning (RAFT). Within the retrieval pipeline, we compare fixed-size sliding-window chunking against structure-aware chunking. The SFT configuration with fixed-size chunking RAG met the criterion on 8 of the 14 examinations, outperforming all alternatives, whereas no configuration without fine-tuning passed any. Aggregate accuracy reached 79.7%, with a confidence interval spanning the threshold, and 80.2% on PWR items specifically. Furthermore, two regularities emerged: the preferred chunking strategy reverses depending on the model’s training state, and RAFT underperforms compared to standard SFT in matching search environments. These results demonstrate which combination of fine-tuning and search approaches achieves operator-level capabilities.
[NLP-14] Scaling Native Multimodal Pre-Training From Scratch
【速读】： 该论文旨在解决生成式 AI（Generative AI）在多模态基础模型训练中缺乏系统性扩展规律的问题，尤其关注在固定计算预算下如何最优分配模型规模与数据 token 数量。其核心挑战在于：尽管原生多模态预训练能够实现视觉与语言之间的深层跨模态融合，并克服传统晚期融合架构中的优化不对称性，但该范式下的模型缩放规律尚未被充分理解。论文的关键解决方案是通过实证研究揭示了在固定计算预算约束下，最小化目标损失遵循可预测的计算定律（compute law），而最优模型规模与 token 数量则呈现幂律缩放关系。进一步发现，语言任务与多模态任务具有显著不同的缩放行为：语言学习的资源分配法则对数据构成高度鲁棒，而多模态任务的分配法则强烈依赖于文本与视觉数据的比例。当文本占比高时，仅在较大模型规模下才能实现计算效率最优，导致资源分配向更高模型容量倾斜。基于此，论文构建了受数据构成影响的计算定律与分配指数模型，进而推导出精确的效率前沿（efficiency frontier），明确指出了模型规模、token 数量与数据混合比例的最佳配置。下游评估结果表明，原生多模态预训练能促进正向跨模态迁移，显著提升纯文本空间推理能力，并支持稳健的多模态上下文学习。综上，本研究为可预测地扩展多模态基础模型提供了关键的实证基础与理论框架。
链接: https://arxiv.org/abs/2607.22043
作者: Haoyuan Wu,Aoqi Wu,Hai Wang,Jiajia Wu,Jinxiang Ou,Bei Yu
机构: The Chinese University of Hong Kong(香港中文大学); LLM Department, Tencent(腾讯大模型部门)
类目: Computation and Language (cs.CL); Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Although large language models (LLMs) exhibit remarkable reasoning capabilities, their reliance on text-only pre-training restricts the perception of the multimodal physical world. Native multimodal pre-training avoids this limitation by training models from scratch on multimodal inputs, thereby achieving deep cross-modal integration and mitigating optimization asymmetries inherent to traditional late-fusion architectures. Despite these advantages, the scaling properties of this paradigm remain systematically uncharacterized. To address this gap, we investigate the optimal model size and token count for training a transformer-based vision-language model under a fixed computational budget. We demonstrate that minimal objective loss adheres to a predictable compute law, whereas compute-optimal model sizes and token counts scale as power laws. Notably, language and multimodal objectives manifest distinct scaling behaviors. The language allocation law is largely invariant to the composition of the data, indicating stable language learning regardless of the multimodal data ratio. Conversely, the multimodal allocation law is highly sensitive to this composition. Specifically, text-heavy mixtures become compute-efficient only at larger model scales, shifting the optimal resource allocation toward greater model capacity. Additionally, by modeling the influence of data composition on compute laws and allocation exponents, we derive an efficiency frontier specifying precise configurations of model size, token count, and data mixture. Downstream evaluations further reveal that native multimodal pre-training induces positive cross-modal transfer, thereby enhancing pure-text spatial reasoning and enabling robust multimodal in-context learning. In summary, this empirical research establishes the essential groundwork for predictably scaling multimodal foundation models.
[NLP-15] Developing and Validating the Spanish Version of the Large Language Models Dependency Scale (LLM -D12-SP)
【速读】： 该论文旨在解决西班牙语群体在日益广泛应用大语言模型（Large Language Models, LLMs）背景下，缺乏可靠且文化验证的心理依赖评估工具的问题。随着LLMs在任务执行、决策支持及组织沟通中的渗透，其引发的心理依赖现象亟需科学测量手段。研究的关键在于首次完成大语言模型依赖量表（LLM-D12-SP）的跨语言心理测量学验证，确认其二维结构——工具性依赖（对LLMs完成任务与辅助决策的依赖）与关系性依赖（对LLMs作为陪伴和社交互动对象的心理依赖）——在西班牙语人群中的适切性。通过386名西班牙语参与者的实证数据，验证了量表具有良好的内部一致性（总α = 0.89）、区分效度以及外部效度，表明两个维度为既相关又独立的心理构念，并与网络成瘾倾向及对LLMs的信任感显著正相关，而与认知需求无关。该研究为组织情境中评估用户对LLMs的心理依赖提供了首个经验证的西班牙语测量工具，拓展了该量表在多语言环境下的普适性与应用价值。
链接: https://arxiv.org/abs/2607.22041
作者: Tran Gia Bao,Mo El-Haj,Sameha Al-Shakhsi,Antonio Garcia-Cabot,Raian Ali,Ala Yankouskaya
机构: 未知
类目: Computation and Language (cs.CL); Computers and Society (cs.CY)
备注: 32 pages
Abstract:There is a growing need for reliable and culturally validated instruments to assess psychological dependency on large language models (LLMs), particularly as LLMs are increasingly used for task execution, decision-making, and communication in organizational and work-related settings. This need is especially relevant for Spanish-speaking populations, where LLM adoption is rapidly expanding, yet validated psychometric tools remain scarce. The present study reports the first validation of the Spanish version of the Large Language Model Dependency Scale (LLM-D12-SP), extending prior validations conducted in English- and Arabic-speaking samples. The LLM-D12 is a two-dimensional instrument assessing Instrumental Dependency (reliance on LLMs for performing tasks and supporting decisions) and Relationship Dependency (psychological reliance on LLMs for companionship and social interaction). A total of 386 Spanish-speaking participants (M = 28.0 years, SD = 6.1; 55% male) completed the LLM-D12-SP. Confirmatory factor analysis supported the original two-factor structure. The scale demonstrated good internal consistency (Cronbach’s alpha = 0.89 total; 0.86 Instrumental; 0.85 Relationship). Discriminant validity analyses indicated that the two subscales represent related but distinct constructs. External validation showed that both dependency dimensions were positively associated with internet addiction and perceived trustworthiness of LLMs, while showing weak or no association with need for cognition. Together with prior English and Arabic validations, these findings establish cross-linguistic support for the scale’s structure and provide a psychometrically sound tool for investigating psychological aspects of LLM use in organizational contexts.
[NLP-16] Enough is as good as a feast: A Comprehensive Analysis of How Reinforcement Learning Mitigates Task Conflicts in LLM s ICLR2026
【速读】： 该论文旨在解决大规模语言模型（LLM）在合并多个专用模型为单一统一模型时所面临的性能退化与任务冲突问题。尽管现有研究多聚焦于优化合并策略，但训练范式（如监督微调SFT与强化学习RL）对模型合并效果的影响尚未得到充分探讨。本文系统比较了基于强化学习（RL）训练的LLM与传统监督微调（SFT）模型在合并过程中的表现，发现RL训练显著降低了任务间的冲突，并在合并后表现出更小的性能下降，因而更适用于模型合并。其解决方案的关键在于揭示了强化学习在模型合并中优越性的三大核心机制：（1）基于策略的训练数据使梯度更新幅度更小，降低了对其他任务已有知识的覆盖风险；（2）强化学习的目标函数“足够即佳”（enough is as good as a feast）在模型收敛过程中逐步减少冲突参数更新的幅度与数量，从而抑制过拟合与参数干扰；（3）强化学习联合优化正负样本，引导模型进入一个无偏的任务特定参数子空间，有效保障各任务性能并进一步缓解参数冲突。
链接: https://arxiv.org/abs/2607.22039
作者: Zixuan Ren,Jinliang Lu,Junhong Wu,Yang Zhao,Dai Dai,Hua Wu,Haifeng Wang,Chengqing Zong
机构: State Key Laboratory of Multimodal Artificial Intelligence Systems, Institute of Automation, CAS(中国科学院自动化研究所多模态人工智能系统重点实验室); School of Artificial Intelligence, University of Chinese Academy of Sciences(中国科学院大学人工智能学院); Baidu Inc.(百度公司)
类目: Computation and Language (cs.CL)
备注: Published in ICLR 2026
Abstract:Model merging plays a crucial role in consolidating multiple specialized models into a single, unified model, especially in the era of large language models (LLMs). Recent research has primarily focused on developing strategies to enhance merging performance with the trained models, while the impact of training paradigms, such as supervised fine-tuning (SFT) and reinforcement learning (RL), on the effectiveness of model merging remains underexplored. In this study, we systematically explore the merging behavior of RL-trained LLMs compared to those trained with traditional SFT. Through comprehensive evaluations across five representative tasks, we find that RL significantly reduces task conflicts and results in less performance degradation after merging, making RL-trained models particularly well-suited for this process. To unearth the reasons behind the superior suitability of RL for model merging, we conduct extensive empirical experiments and theoretical analyses. Our findings highlight three key factors: (1) On-policy training data in RL control the gradient updates in a smaller magnitude, reducing the risk of overwriting existing knowledge for other tasks in the model. (2) The RL optimization objective, which favors ``\textitenough is as good as a feast", progressively reduces the magnitude and the number of conflict parameter updates as the model converges. (3) Joint optimization of positive and negative examples in RL steers the model towards an unbiased task-specific parameter subspace, ensuring robust performance while further preventing parameter conflicts.
[NLP-17] Small Vision-Language Models Know When They Are Wrong But Cannot Say So: A Two-Model Study of Stated versus Internal Confidence Under Realistic Image Degradation
【速读】： 该论文旨在解决在消费级硬件部署场景下，视觉语言模型（Vision-Language Models, VLMs）面对真实世界图像退化（如压缩、相机抖动、光照不良）时，如何可靠地提供不确定性信号的问题。在这些条件下，系统的决策可靠性依赖于能否准确识别自身置信度不足的场景并选择延迟响应，而非盲目输出结果。研究的关键发现是：尽管模型在自然语言中表达的“口头置信度”表现出高度稳定但缺乏区分能力（如Qwen2-VL在各种退化条件下平均置信度维持在0.87–0.90，错误检测性能接近随机水平，AUROC≈0.50），而模型内部生成答案的“平均词元概率”则能显著区分正确与错误预测（AUROC达0.92–0.99），展现出更强的误差检测能力。此外，在小规模模型SmolVLM中，口头置信度几乎无法有效提取，而内部概率仍保持较高判别性。然而，当面临严重欠曝等极端低光条件时，两种置信信号均失效，模型性能急剧下降且置信度变化微弱，导致无法可信地触发延迟机制。因此，解决方案的核心在于：利用模型内部的自知识（self-knowledge）——即基于生成过程的内部概率分布作为更可靠的不确定性信号，而非依赖其自然语言输出中的口头置信度；同时强调，在严重光照不足条件下，任何置信信号均不可靠，应避免依赖。
链接: https://arxiv.org/abs/2607.22034
作者: M M Asif Ferdous
机构: Independent Researcher(独立研究员)
类目: Computer Vision and Pattern Recognition (cs.CV); Computation and Language (cs.CL); Machine Learning (cs.LG)
备注: 15 pages, 4 figures
Abstract:Vision-language models (VLMs) are increasingly deployed on consumer hardware where input images are degraded by compression, camera shake, and poor lighting. In such settings, a reliable uncertainty signal matters more than raw accuracy, because it determines when a system should defer rather than answer. We evaluate two small open-weight VLMs – Qwen2-VL-2B-Instruct and SmolVLM-Instruct – across six realistic photographic degradations at three severity levels, comparing two confidence signals: the confidence the model states in natural language, and the model’s own mean token probability over its generated answer. Across 3,800 predictions, we find a large and consistent gap. Verbalized confidence in Qwen2-VL is almost constant (mean 0.87-0.90 across all conditions) and detects its own errors at chance level (AUROC 0.39-0.75, typically ~0.50), while internal token probability from the same model separates correct from incorrect answers with AUROC 0.92-0.99. In SmolVLM, verbalized confidence proved largely unobtainable: across three prompt templates, only one of five pilot attempts produced a parseable confidence value, while internal probability again yielded above-chance error detection (AUROC 0.54-0.92). Both models fail in the same place: under severe underexposure, accuracy collapses (0.99-0.22 for Qwen2-VL, 0.97-0.42 for SmolVLM) while both confidence signals barely move, and internal error-detection falls to chance. We conclude that small VLMs encode usable self-knowledge that their verbalized output does not express, that internal probability is therefore the better deferral signal in constrained deployment, and that neither signal should be trusted under severe low-light conditions.
[NLP-18] DWT-Fusion: A Signal-Based Framework for Training-Free LLM -Generated Text Detection
【速读】： 该论文旨在解决在零样本（zero-shot）和无训练（training-free）条件下，检测由大语言模型（LLM）生成文本的难题，尤其关注检测器在跨数据集、跨领域及面对未见过的生成器时的泛化能力。现有无训练方法通常依赖于全局统计量来表征文本的语言模型行为，忽略了词元级可预测性中潜在的局部与多尺度变化信息。为此，本文提出DWT-Fusion框架，其核心创新在于利用代理因果语言模型生成的词元级对数概率序列，通过离散小波变换（Discrete Wavelet Transform, DWT）进行多分辨率信号分析，从而捕捉局部概率动态变化，提取更具判别性的检测信号。该框架进一步引入四种无训练投票机制（包括等权重硬投票、等权重软投票、校准加权硬投票与校准加权软投票），在不训练监督型元分类器的前提下融合多个小波配置结果。实验在HC3、M4和MAGE数据集上使用GPT-Neo-2.7B、GPT-J-6B、Falcon-7B和LLaMA-3-8B作为代理模型进行评估，结果显示，最优单个波形配置在三个数据集上的AUROC分别达到0.9872、0.8185和0.7138；经校准加权投票融合后，性能进一步提升至0.9919、0.8477和0.7471。研究证明，基于小波的多分辨率评分与校准引导的投票融合策略能够提供高效且可解释的无训练检测信号，显著提升了对不同来源和场景下生成文本的识别能力。
链接: https://arxiv.org/abs/2607.22026
作者: Mehmet Batuhan Özdaş,Murat Osmanoğlu
机构: Ankara University (安卡拉大学); Ankara University (安卡拉大学)
类目: Computation and Language (cs.CL)
备注: 40 pages, 4 figures, 6 tables
Abstract:Detecting LLM-generated text remains challenging under zero-shot and training-free conditions, especially when detectors must generalize across datasets, domains, and unseen generators. While existing training-free approaches exploit language-model statistics as detection signals, they typically characterize a text through global measures that summarize overall model behavior. Consequently, potentially informative local and multiscale variations in token-level predictability may remain underutilized. Motivated by this observation, we introduce DWT-Fusion, a training-free signal-based framework for detecting LLM-generated text using discrete wavelet analysis of token-level log-probability sequences produced by a proxy causal language model. The proposed framework analyzes these sequences through wavelet-based multiresolution signal representations and derives detection signals from localized probability dynamics. We further evaluate four training-free voting variants, including equal-weight hard voting, equal-weight soft voting, calibration-weighted hard voting, and calibration-weighted soft voting, to combine multiple wavelet configurations without training a supervised meta-classifier. We evaluate the framework on HC3, M4, and MAGE using GPT-Neo-2.7B, GPT-J-6B, Falcon-7B, and LLaMA-3-8B as proxy models. The best single wavelet configurations achieve AUROC values of 0.9872, 0.8185, and 0.7138 on HC3, M4, and MAGE, respectively. With calibration-weighted voting, the best ensemble variants further improve AUROC to 0.9919, 0.8477, and 0.7471. These findings show that DWT-based multiresolution scoring and calibration-guided voting fusion provide effective and interpretable signals for training-free LLM-generated text detection.
[NLP-19] Zero-Shot Mission-Level Evaluation for Aerial MLLM Agents
【速读】： 该论文旨在解决多模态大语言模型（Multimodal Large Language Models, MLLMs）在具身智能体执行长时序、高阶任务时的通用性与能力边界问题，尤其关注模型能否仅凭单一高层指令在复杂三维空中环境中完成多步骤自主决策。其核心挑战在于评估通用模型在无特定领域微调的情况下，是否具备跨环境、跨任务的零样本（zero-shot）具身推理能力。解决方案的关键是提出MissionBench——一个面向空中三维环境的任务级评估基准，涵盖5个模拟3D环境、120项任务及4类任务类型，要求智能体基于第一人称观测和动作历史进行自主规划、导航与结果报告。实验表明，当前最强的22个开源与闭源MLLM在任务成功率上普遍低于35%，远低于人类84.4%的表现，揭示了现有模型在多步规划、自适应推理等综合能力上的显著不足。研究进一步发现，尽管不同模型家族间表现差异大，但模型规模的扩大仍带来性能提升，说明更大的通用模型具备更强的零样本具身能力。分析强调，任务级成功不仅依赖空间感知，更需协调多步规划与动态推理等多重能力，从而推动闭环评估范式的发展，并凸显了基于规模扩展提升具身人工智能潜力的同时伴随的风险。
链接: https://arxiv.org/abs/2607.22014
作者: Suman Navaratnarajah,Taehyoung Kim,Jona Ruthardt,Ishaan Bhimwal,Ryousuke Yamada,Yannik Blei,Wolfram Burgard,Yuki M Asano
机构: 未知
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Computer Vision and Pattern Recognition (cs.CV); Robotics (cs.RO)
备注: Preprint
Abstract:Multimodal Large Language Models (MLLMs) are emerging as core reasoning modules for embodied agents, yet it remains unclear how well general-purpose models can solve long-horizon embodied tasks from a single high-level instruction. We introduce MissionBench, a benchmark for mission-level evaluation of MLLMs in aerial 3D environments. It comprises 120 missions across five simulated 3D environments and four task families. Agents must autonomously plan, navigate, and report outcomes using only egocentric observations and its action history, without aerial-specific fine-tuning. Across 22 open- and closed-source MLLMs, the strongest model succeeds on fewer than 35% of missions compared to 84.4% human performance, highlighting the difficulty of multi-step embodied tasks. Despite large variations between model families, we observe gains from scaling, indicating that larger general-purpose models possess stronger zero-shot embodied capabilities. Our analysis shows that mission-level competence requires coordinating multiple capabilities beyond spatial perception, including multi-step planning and adaptive reasoning. This motivates closed-loop evaluation and highlights both the promise and risk of scaling-driven improvements for embodied AI.
[NLP-20] Analysing Self-Harm Representations in Language Models: a Cross-Architecture Study
【速读】： 该论文旨在解决生成式人工智能（Generative AI）在识别自残内容时面临的高难度自然语言处理（NLP）挑战，这一任务具有极高的实际意义，需确保极高准确率以实现对高风险用户的及时干预或标记。其核心解决方案的关键在于深入分析大语言模型（LLM）对自残内容的内部表征机制。研究聚焦于两个自残文本数据集（X-Sensitive 和 SH-Detection）及四款主流模型，通过两项关键实验揭示：首先，在所有模型中，自残信息的语义表征主要集中于网络层深度的最后3%至7%（即93%至97%层深），表明高层特征对自残意图具有显著敏感性；其次，通过提取对比性的自残方向并进行归一化处理，发现最优探测器并非总是线性可分性最强的，尤其在Gemma-3-4B模型中观察到其以更复杂、非线性的方式表征自残语义，暗示不同模型在安全表征学习上存在本质差异。该发现为自残内容检测、模型干预策略设计及内容治理提供了重要的理论依据与技术路径。
链接: https://arxiv.org/abs/2607.21988
作者: Luis Espinosa-Anke,Carla Perez-Almendros
机构: Cardiff University (卡迪夫大学)
类目: Computation and Language (cs.CL)
备注:
Abstract:Self-harm content is particularly challenging to detect using NLP techniques, and is also a high-stakes task which requires the highest accuracy to enable timely intervention or flagging at-risk users. We therefore present an analysis of how LLMs represent such self-harm content, which has downstream applications in self-harm detection, LLM intervention and governance and policing. In this paper, we focus on two datasets and four models, and perform two main experiments: (1) We train and evaluate linear probes across all layers of each model on two self-harm datasets: X-Sensitive and SH-Detection. Across both corpora, self-harm information crystallizes in the final 3 - 7% of network layers (93 to 97% depth). (2) We extract contrastive self-harm directions and, after performing a normaliation step, we find that the most accurate probes are not necessarily the most linearly separable. In particular, we find Gemma-3-4B to represent this \textitcontrastive self-harm direction in a slightly different, more intricate way than the other LLMs.
[NLP-21] J-CoT: Chain-of-Thought in J-Space
【速读】： 该论文旨在解决生成式语言模型在链式思维（Chain-of-Thought, CoT）推理过程中依赖自然语言作为唯一可传递状态接口所带来的局限性。现有方法虽通过隐式推理（Latent Reasoning）允许连续隐藏状态的递归传播，但其采用稠密隐藏向量整体传递信息，缺乏对下一推理步骤所需关键信息的显式选择与组织机制。为此，本文提出J-CoT，一种基于J空间（J-space）——即模型隐藏表示中以词汇索引为坐标的坐标系统——的递归推理框架。J-CoT在每个推理周期内于完整隐藏空间中计算，并在周期边界将中间状态表达为词汇索引系数，以“J-思考”（J-thought）形式向前传递，再映射回模型隐藏表示以进行下一轮推理。该方法既保持了语言学上的可解释性，又避免了对完整解码语句的需求，实现了信息选择与结构化传递的显式控制。实验表明，在相同骨干网络与推理设置下，J-CoT-Zero在所有基准测试中表现达到或超越最强的隐式推理基线；而J-CoT-Train在数学、科学、编程及结构化路径推理等任务上均取得最高得分，验证了其高效的信息组织能力与泛化性能。
链接: https://arxiv.org/abs/2607.21981
作者: Junde Wu,Jiayuan Zhu,Fengling Liu,Minhao Hu,Jiazhen Pan
机构: University of Oxford (牛津大学); Imprint Lab (印迹实验室); Stanford University (斯坦福大学)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: work in progress
Abstract:Chain-of-thought prompting improves language-model reasoning by carrying intermediate states across successive computation steps. However, relying on natural language as the only recurrent interface is overly restrictive, since many transient computations do not need to be fully verbalized. Existing latent-reasoning methods remove this constraint by recurrently propagating continuous hidden states. However, these methods pass a dense hidden vector as a whole, without an explicit mechanism for selecting and organizing the information needed by the next reasoning step. This motivates an intermediate interface that remains linguistically grounded without requiring a decoded sentence. We introduce \textbfJ-CoT, a recurrent reasoning framework built on \emphJ-space, a vocabulary-indexed coordinate system within the model’s hidden representations. Within each cycle, the model computes in its full hidden space. At the cycle boundary, J-CoT expresses the intermediate state as vocabulary-indexed coefficients, carries these coefficients forward as a \emphJ-thought, and maps them back into the model’s hidden representation for the next cycle. J-CoT therefore requires neither a fluent intermediate rationale nor recurrence over the complete hidden state. Under matched backbone and inference settings, J-CoT-Zero matches or exceeds the strongest evaluated latent-reasoning baseline on every benchmark, while J-CoT-Train obtains the highest score across the evaluated mathematical, scientific, coding, and structured path-reasoning tasks.
[NLP-22] Analyzing Toxic Behavior and Its Impact on the Mastodon Community
【速读】： 该论文旨在解决去中心化社交平台Mastodon中毒性内容（toxic content）检测与治理的难题。由于其由独立管理的社交服务器组成的联邦架构，缺乏统一的审核标准且生态多样性与治理能力参差不齐，导致毒性内容的识别与干预面临显著挑战。论文的关键解决方案在于运用机器学习方法对用户发布内容进行分析，以揭示毒性内容在平台内的演化模式与传播特征，从而为社区健康维护与去中心化治理机制优化提供数据驱动的洞见。
链接: https://arxiv.org/abs/2607.21980
作者: Pasan Kamburugamuwa,Scrivner,Olga B
机构: Indiana University Bloomington (印第安纳大学布卢明顿分校)
类目: Computation and Language (cs.CL); Computational Engineering, Finance, and Science (cs.CE)
备注: 5 pages, 1 figure
Abstract:Mastodon as a decentralized federation of independently moderated social servers poses unique challenges for the detection and mitigation of toxic content. There are no unified moderation standards. The ecosystem is very diverse and uneven. This paper explores the development and spread of toxicity in Mastodon, utilizing machine learning methods to examine user posts. The results offer clarity on toxicity trends and its implications for community health and decentralized governance.
[NLP-23] MoE2-LoRA: When MoE Models Meet MoE-style Low-Rank Adaptation
【速读】： 该论文旨在解决大语言模型中混合专家（Mixture-of-Experts, MoE）架构在参数高效微调（Parameter-Efficient Fine-Tuning, PEFT）方面存在的关键问题：现有方法或忽略路由（router）先验信息而采用均匀适配器，导致效率低下并存在遗忘风险；或依赖静态专家选择，限制了逐令牌的容量表达能力及跨专家特征学习。其核心解决方案是提出一种新型的MoE风格低秩适应方法——MoE²-LoRA，通过双通道路由条件投影（Routing-Conditioned Projection, RCP）模块，将预训练专家的专业化特性与任务特定的可适应性深度耦合，复用基础路由激活以指导LoRA的路由决策。此外，引入一个跨所有层共享的全局LoRA专家池，实现了全模型范围的自适应与涌现的分层亲和性及专家利用均衡。该方法同时兼顾了先验知识复用、动态适配路由与全模型知识共享的优势，在多种不同规模与专家粒度的MoE骨干网络上均实现最优下游性能，并保持更强的泛化能力。
链接: https://arxiv.org/abs/2607.21978
作者: Qingyu Yang,Haonan He,Minglei Li,Jingqi Ye,Tao Chen,Lei Bai,Peng Ye
机构: Shanghai Artificial Intelligence Laboratory; KTH Royal Institute of Technology; University of Science and Technology of China; Fudan University; The Chinese University of Hong Kong
类目: Computation and Language (cs.CL)
备注: Preprint, under review
Abstract:Mixture-of-Experts (MoE) architectures have been widely adopted in large language models, yet parameter-efficient fine-tuning (PEFT) for MoE models remains underexplored. Existing PEFT methods for MoE either ignore router priors with uniform adapters, reducing efficiency and risking forgetting, or rely on static expert selection, limiting per-token capacity and cross-expert feature learning. In this paper, we make the first attempt to fine-tune MoE models with MoE-style low-rank adaptation: our method, entitled MoE ^2 -LoRA, deeply couples the pretrained expert specialization with task-specific adaptivity via a dual-channel Routing-Conditioned Projection (RCP) module, which reuses base router activations to inform LoRA routing. We further introduce a single global LoRA expert pool shared across all layers, enabling model-wide adaptation with emergent layer-wise affinities and balanced expert utilization. MoE ^2 -LoRA simultaneously benefits from the advantages of prior reuse, dynamic adapter routing, and model-wide knowledge sharing. Evaluated on multiple MoE backbones with varying scales and expert granularities, MoE ^2 -LoRA consistently achieves state-of-the-art downstream accuracy while retaining stronger general capabilities.
[NLP-24] aching LLM s to Self-Evolve: Cultivating Core Meta-Skills with Reinforcement Learning
【速读】： 该论文旨在解决传统后训练方法在提升模型自我进化能力方面的局限性，特别是其未能有效培养如“基于环境反馈的自我反思”等元技能（meta-skills），而这些技能对于实现多轮迭代优化至关重要。为弥补这一差距，论文提出MetaEvolve框架，其核心解决方案在于通过数据合成管道、面向进化的强化学习（evolution-aware reinforcement learning, RL）以及推理时的演化搜索，系统性地构建和训练模型的自我进化元技能。具体而言，该框架以编程任务为应用领域，利用程序执行结果提供连续的奖励信号（超越传统的正确性二值判断），并基于此构建包含当前代码、适应度评分（融合正确性与效率）及历史尝试记录的演化轨迹作为训练数据，采用可验证的奖励信号进行强化学习训练。通过大规模代码数据训练，旨在习得可泛化的、跨领域的通用元技能，从而在缺乏丰富训练信号的开放性问题中仍具备强大适应能力。实验表明，MetaEvolve在7个编码基准上相较最强基线在分布内任务上提升10.01%，分布外任务上提升24.12%，并在完全脱离训练域的开放算法优化问题中实现46.9%的相对改进，证明了显式培养自我演化元技能是迈向更强大、自主演化的智能体的可行且有效的路径。
链接: https://arxiv.org/abs/2607.21971
作者: Shujin Wu,Cheng Qian,Xiusi Chen,Heng Ji
机构: University of Illinois Urbana-Champaign (伊利诺伊大学厄本那-香槟分校)
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注:
Abstract:Test-time scaling through iterative self-evolution with environment feedback, as demonstrated by AlphaEvolve, shows remarkable performance gains. We hypothesize that the success of such evolution frameworks hinges on meta-skills, such as self-reflection with environment feedback, that enable effective multi-round refinement, yet are largely neglected by traditional post-training. To bridge this gap, we present MetaEvolve, a framework designed to develop these meta-skills via a data synthesis pipeline, evolution-aware reinforcement learning (RL), and inference-time evolutionary search. Concretely, we ground MetaEvolve in coding, where program execution provides natural, continuous reward signals beyond binary correctness. Building on these signals, we synthesize evolution trajectories as training data, each containing a current program, its fitness score (combining correctness and efficiency), and a history of prior attempts, and train the model via RL with verifiable rewards derived from test case execution. By training on large-scale code data, we aim to inspire generalizable domain-agnostic meta-skills that can transfer broadly to open-ended problems where such rich training signals are scarce. Across seven coding benchmarks, MetaEvolve outperforms the strongest baseline by 10.01% absolute on in-distribution tasks and 24.12% on out-of-distribution tasks. On open-ended algorithm optimization problems entirely outside the training domain, it further achieves a 46.9% relative improvement. These results demonstrate that explicitly cultivating self-evolution meta-skills offers a principled path toward more capable and autonomously self-evolving AI.
[NLP-25] Ground Truth First: A Longitudinal Evaluation Instrument for Agent Memory and the Tenure Crossover in Memory-Architecture Rankings
【速读】： 该论文旨在解决现有大语言模型（LLM）代理记忆评估基准中存在的标签错误（label-error）、数据污染（contamination）以及过度聚焦短期交互历史等问题。其核心解决方案在于重构评估流程：采用“种子生命脚本采样器”在文本生成前预先定义事实的生效区间、波动性类别及信息来源渠道；随后由大语言模型（LLM）渲染聊天与邮件内容，基于事件级事实清单生成对话；通过保真度验证器确保所有预植事实均被正确保留；最终问题由脚本机械生成，保证黄金答案（gold answers）在脚本层面有效且独立验证可回答性。所构建的合成虚构语料库（约380个问题，15种类型）引入了多项基准中缺失的关键特征，包括事实的有效期、发送/接收信任区分、良性环境下的注入探测样本，以及“截至日期”问题集。在对五种记忆架构与无记忆对照组（固定回答者、版本化LLM裁判、三重复制、两个时间跨度）进行基准测试时发现，随着历史长度增加，后端排名发生反转：预算受限的精选地图记忆在三周时表现最优，但在九周时因淘汰内容召回率从96%降至72%，而溯源类型图结构则提升至90%；该反转在六名用户全部跨家族重新评判下仍显著（精确p=0.031）。全历史渲染基线在短期表现上优于或持平最佳记忆系统，但在九周时未展现裁判无关优势，且读取成本约为两倍。写入阶段质量与下游表现强相关（弱写事实失败率为24% vs 2%），且注入抗性取决于溯源边界是否在表征中得以保留。综合来看，分层架构在两种时间尺度下均表现最佳（短期96.8%），并已作为开源库Veracium发布，包含语料生成器与评估框架。
链接: https://arxiv.org/abs/2607.21962
作者: Quentin Spencer
机构: 未知
类目: Computation and Language (cs.CL)
备注: 25 pages, 2 figures. Code: this https URL
Abstract:Benchmarks for LLM-agent memory typically generate conversations first and extract answer keys afterwards – with documented label-error and contamination problems – and they overwhelmingly measure short interaction histories. We invert the pipeline: a seeded life-script sampler emits facts with validity intervals, volatility classes, and source channels before any text exists; an LLM renderer writes chat and email from per-event fact manifests; a fidelity verifier confirms every planted fact; and questions are instantiated mechanically from the script, so gold answers are script-valid by construction and separately validated for answerability. The synthetic, fictionalized corpus (~380 questions, 15 types) embeds features absent from the benchmarks we survey: per-fact validity intervals, sent/received trust distinctions, injection probes in a benign harness, and as-of-date question sets. Benchmarking five memory architectures against a no-memory control (fixed answerer, versioned LLM judge, three replicates, two horizons), we find backend rankings invert with history length: the budgeted curated-map memory that leads at three weeks loses recall of evicted content by nine weeks (96% to 72%) while a provenance-typed graph rises to 90%; the inversion is positive for all six users under complete cross-family re-judging (exact p=0.031). A full-rendered-history baseline ties or exceeds the best memory system at the short horizon but shows no judge-independent advantage at nine weeks, at about twice the read cost. Write-stage quality strongly correlates with downstream quality (weakly-written facts fail 24% vs 2%), and injection resistance tracked whether provenance boundaries survive representation. A layered architecture performs best among the memory systems in both regimes (96.8% short-horizon) and is released as Veracium, an open-source library, with the corpus generator and harness.
[NLP-26] On Improving Faithfulness of Podcasts from Documents
【速读】： 该论文旨在解决生成式对话内容（如播客）在长篇多角色对话中缺乏事实一致性的问题，即模型生成的内容与源文档之间存在未被支持的、虚构的信息（ungrounded information）。其核心挑战在于如何在跨多轮对话的复杂语境下维持生成内容对源文档的忠实性（faithfulness）。解决方案的关键是提出一种模型无关的“catch-n-repair”框架，该框架通过一个逐轮评估的大型语言模型作为裁判（LLM-as-a-judge）机制，识别出不忠实的对话回合，并对其进行修复，同时保持对话的连贯性和自然流畅性。实验表明，该方法在域内和域外设置下均显著提升了生成内容的忠实度，有效缓解了当前先进模型（如GPT-4o）普遍存在的幻觉问题。
链接: https://arxiv.org/abs/2607.21961
作者: Soumya Dutta,Tejas Indulal Dhamecha,Pannaga Shivaswamy
机构: Adobe Research(Adobe 研究院)
类目: Computation and Language (cs.CL)
备注: Under Submission
Abstract:Large language models (LLMs) are increasingly used to generate long-form conversational content such as podcasts from textual sources. While these systems produce fluent and engaging narratives, they often introduce ungrounded information. In this work, we present the first systematic study of faithfulness in document-grounded podcast generation, where grounding must be maintained across conversational turns in long-form, multi-speaker transcripts. We construct a dataset of over 1500 documents spanning five domains and generate podcast transcripts using multiple LLMs. We introduce a turn-level LLM-as-a-judge framework for evaluating whether conversational turns are supported by the source document, and validate its reliability through human studies. Our analysis shows that even state-of-the-art models, including GPT-4o, frequently generate ungrounded content. To mitigate this issue, we propose catch-n-repair, a model-agnostic framework that detects and rewrites unfaithful conversational turns while preserving conversational flow. Experiments demonstrate consistent improvements in faithfulness across both in-domain and out-of-domain settings.
[NLP-27] Leverag ing External Knowledge for Historical Document Restoration via Retrieval-Augmented Large Language Models ACL2026
【速读】： 该论文旨在解决历史文献因物理退化导致的字迹模糊或缺失问题，尤其针对现有基于掩码语言建模（masked language modeling）的方法在恢复需要外部历史知识支撑的专有名词（named entities）时表现不足的局限性。其解决方案的关键在于提出一种基于检索增强生成（retrieval-augmented generation, RAG）的大语言模型框架——ARI，通过融合预训练大模型的隐式知识与显式检索获取的外部历史上下文，有效提升对依赖上下文的专有名词的推断能力。实验结果表明，该方法在韩语历史文献上的恢复性能显著优于基线模型，尤其在通用字符和专有名词的恢复上均取得显著提升，且专家评估进一步验证了其在实际历史研究中的可用性与价值。
链接: https://arxiv.org/abs/2607.21936
作者: Gabeen Kim,Kyeongpil Kang
机构: Kangwon National University (江原国立大学)
类目: Computation and Language (cs.CL)
备注: Accepted to Findings of ACL 2026
Abstract:Historical documents act as invaluable knowledge archives but often suffer from illegibility due to physical deterioration and damage. While existing restoration methods based on masked language modeling effectively utilize local context, they struggle to restore named entities that require external historical knowledge. To address this limitation, we introduce a novel framework for historical document restoration that leverages large language models with retrieval-augmented generation (RAG). By combining the implicit knowledge of pre-trained LLMs with explicitly retrieved external context, our model ARI effectively mitigates the challenge of inferring context-dependent proper nouns. Extensive experiments on Korean historical documents demonstrate that our approach significantly outperforms baselines, achieving substantial gains in restoring both general characters and named entities. Furthermore, comprehensive evaluations including expert assessments confirm that ARI serves as a practical tool for domain experts, promising to accelerate the analysis of historical records.
[NLP-28] Diffusion Models in Medical Image Inpainting: Challenges Solution Taxonomy and Future Directions
【速读】： 该论文旨在解决医学图像中因伪影、信息缺失或病理性改变导致的视觉与语义不一致问题，进而影响诊断可靠性及下游临床应用。其核心挑战在于如何在保持解剖结构合理性的同时，精准重建受损区域。解决方案的关键在于采用基于扩散模型（diffusion models）的生成式方法，尤其是去噪扩散概率模型（denoising diffusion probabilistic models）和潜空间扩散模型（latent diffusion models），这些模型能够生成具有解剖学一致性的高质量图像重建结果。通过系统性综述60项相关研究，论文构建了针对扩散模型在医学图像修复中的分类体系，并揭示了当前研究主要聚焦于磁共振成像（MRI）与计算机断层扫描（CT）中的伪影去除、数据增强、伪健康组织重建及异常检测等任务。尽管扩散模型在生成合理解剖结构方面表现优异，但其广泛应用仍受限于缺乏标准化评估基准、数据集多样性不足以及跨临床场景验证机制不完善等关键问题。
链接: https://arxiv.org/abs/2607.21904
作者: Arthur Dantas Mangussi,Joana Cristo Santos,Ricardo Cardoso Pereira,Ana Carolina Lorena,Mário A. T. Figueiredo,Pedro Henriques Abreu
机构: Aeronautics Institute of Technology and Science and Technology Institute, Federal University of São Paulo (巴西圣何塞多斯坎普斯航空技术学院及圣保罗州联邦大学科学技术研究所); LASIGE, Faculdade de Ciências, Universidade de Lisboa (里斯本大学科学学院LASIGE实验室); University of Coimbra, CISUC/LASI – Centre for Informatics and Systems of the University of Coimbra, Department of Informatics Engineering (科英布拉大学，科英布拉大学信息与系统中心/科英布拉大学信息工程系); Instituto Superior Técnico, Universidade de Lisboa, Instituto de Telecomunicações (里斯本高等技术学院，里斯本大学，电信研究所)
类目: Computer Vision and Pattern Recognition (cs.CV); Computation and Language (cs.CL)
备注:
Abstract:Image inpainting aims to reconstruct missing or corrupted regions of an image while preserving as much as possible, visual and semantic consistency. In medical imaging, this task is particularly important because artifacts, missing information, and pathological alterations can compromise diagnostic reliability and downstream clinical applications. Recently, diffusion models have emerged as state-of-the-art generative approaches for medical image inpainting due to their ability to generate anatomically consistent reconstructions. This survey presents a systematic review of diffusion-based methods for medical image inpainting, covering the main architectures, applications, datasets, and evaluation strategies reported across 60 studies. In addition, we propose a taxonomy for diffusion-based approaches. The analysis reveals a rapid growth of research interest in diffusion-based medical image inpainting, with denoising diffusion probabilistic models and latent diffusion models emerging as the dominant architectures. The reviewed studies mainly focus on artifact removal, data augmentation, pseudo-healthy tissue reconstruction, and anomaly detection, particularly in magnetic resonance imaging and computed tomography imaging. Overall, diffusion models demonstrate strong performance in producing anatomically plausible reconstructions and aiding downstream clinical tasks. However, the review also highlights important challenges, including the lack of standardized benchmarks, limited dataset diversity, and restricted validation procedures across diverse clinical applications and imaging scenarios.
[NLP-29] Data Quality over Capacity: Internalizing Documents into LoRA Adapters for Closed-Book QA
【速读】： 该论文旨在解决在不依赖外部检索或上下文窗口预算的情况下，使模型能够通过闭卷方式准确回答关于文档语料库的问题。其核心挑战在于如何在有限的参数容量下实现高效的知识内化。解决方案的关键在于利用低秩适配器（LoRA）将文档知识直接“烘焙”至4位量化版Gemma-4-e4b模型的权重中，并通过系统性实验揭示：当适配器容量足够时，训练数据质量是决定闭卷准确率的主导因素，显著超过LoRA秩、学习率及替代架构设计的影响；同时发现适配器容量存在一个硬性阈值，低于该阈值则任何数据干预均无效。研究进一步表明，通过一次简练的数据清洗（将黄金答案精简为1–6词的标准片段并剔除琐碎信息），闭卷准确率在15篇文档语料上从57.7%提升至85.7%，效果远超架构调整。此外，研究还验证了适配器秩需随语料规模增长的趋势，并纠正了早期对秩与学习率之间耦合关系的误判。最终，在15篇文档子集上引入真实检索基线后，内化适配器（84.2%召回率）优于基于BM25-RAG的基线阅读器（58.9%）甚至接近理想分块检索器（65.6%），且延迟更低。本文以完整调试过程（含三次误判）为案例，展示了大语言模型训练中基于实证的调试方法论。
链接: https://arxiv.org/abs/2607.21861
作者: Joan Figuerola Hurtado
机构: 未知
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注:
Abstract:We study baking documents directly into the weights of a 4-bit Gemma-4-e4b model via LoRA, so a system can answer questions about a corpus closed-book: no retrieval and no context-window budget. Across roughly 100 training runs from single documents to a 99-document corpus, we find that once adapter capacity is adequate, training-data quality is the dominant lever on closed-book accuracy, outweighing LoRA rank, learning rate, and two alternative architectures combined; capacity itself is a hard gate below which no data intervention helps. A single curation pass (shortening gold answers to canonical 1-6 word spans and dropping trivia) moved closed-book accuracy from 57.7% to 85.7% on a 15-document corpus, a larger jump than any architectural change. We confirm a capacity trend (rank must grow with corpus size) entangled with a coupling between rank and learning rate that we initially misdiagnosed. On a 15-document slice we add a real retrieval baseline: the internalized adapter (84.2% recall) beats a BM25-RAG pipeline with a base reader (58.9%) and even a realistic gold-chunk oracle (65.6%) at lower latency. We report the full arc, including three misdiagnoses, as a case study in debugging LLM training empirically.
[NLP-30] LeAct: Learning to Reason from Expert Actions
【速读】： 该论文旨在解决生成式模型在推理能力训练中缺乏高质量、可泛化的自然语言推理链（Chain of Thought, CoT）数据的问题。当前主流方法依赖人工标注或从更强的大语言模型（LLM）中蒸馏知识，但这些途径存在成本高、规模有限等局限。论文提出利用专家系统（如游戏引擎、经典规划器、定理证明器）作为新的监督来源，尽管这些系统仅输出最优动作而无显式推理过程，其内部决策逻辑是“沉默”的。核心挑战在于如何从单一动作反推其背后的隐含推理链。为此，论文提出LeAct（Learning to reason from Actions）框架，将推理链视为潜在变量，并通过优化该变量实现逆向重构：学生模型对每个专家动作采样多个候选推理链，仅保留能显著提升自身恢复该动作概率的推理链。实验表明，在多尺度不完全信息博弈（如Flop Hold’em，约10⁹个信息集）和模拟机器人基准任务中，LeAct不仅在可枚举的小规模游戏中逼近求解器的理论最优性能，且在大规模场景下比最强基线提升5倍接近求解器水平；在扑克对局中以+60毫筹码/手的优势胜出，在机器人任务中唯一实现超越直接模仿学习的效果。该研究建立了一个严谨的框架，首次将专家系统转化为大模型可学习的推理教师，为构建具备强泛化能力的推理模型开辟了新路径。
链接: https://arxiv.org/abs/2607.21856
作者: Ziran Yang,Chengshuai Shi,Raj Ghugare,Benjamin Eysenbach,Karthik Narasimhan,Chi Jin
机构: Princeton University (普林斯顿大学)
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注: 27 pages, 3 figures, 11 tables
Abstract:Modern reasoning models depend on reasoning data, today sourced from human annotations or distilled from stronger LLMs. However, a rich and largely untapped source of supervision lies in expert systems (e.g., game engines, classical planners, theorem provers), which routinely produce near-optimal actions across diverse domains. But these experts are silent: they commit to an action without writing down the chain of thought (CoT) behind it. Recovering that CoT as natural-language reasoning would distill expert knowledge into a student that generalizes beyond the demonstrated actions. We treat it as a latent variable and study how to recover it from the action alone. Our approach, LeAct (Learning to reason from Actions), optimizes this latent variable: the student samples candidate CoTs for each expert action, and we retain those that measurably improve its own probability of recovering the action. Across imperfect-information games at multiple scales and a simulated robotics benchmark, LeAct reaches the solver’s numerical floor on small enumerable games. At larger scale, it is 5\times closer to the solver than the strongest expert-iteration baseline. At Flop Hold’em ( \sim 10^9 infosets), LeAct wins head-to-head by +60 mbb/g, and on the robotics probe it is the only training recipe that improves on direct imitation. We present a principled framework and the result: expert systems become a categorically new source of reasoning teachers for foundation models.
[NLP-31] Adversarial Prompts for Acceptance Collapse in Speculative Decoding
【速读】： 该论文旨在解决生成式AI（Generative AI）中无损加速技术（如推测解码，speculative decoding）所面临的安全性问题。尽管推测解码通过动态对齐草稿模型与目标模型的词元（token-level alignment）实现了显著的推理加速，但其语义等价性保证背后隐藏着严重的操作漏洞：攻击者可通过精心设计的提示后缀（prompt-suffix）破坏验证器（verifier）对草稿输出的接受率。为此，论文提出ADSD（Adversarial Suffix for Speculative Decoding），据我们所知，这是首个针对推测解码的提示后缀攻击方法。其核心在于利用“软坍缩”（Soft-Collapse）机制——一种基于非对称推测接受规则构建的、与验证器对齐的代理损失函数，结合目标模型保真度约束（target-preservation objective），在不明显破坏任务质量的前提下，将草稿模型的概率质量推向目标模型难以接受的词元。实验表明，在GSM8K数据集上，该攻击使平均样本时间增加62.3%，同时保持任务输出质量。此外，研究还验证了该漏洞在不同领域、推测解码策略及模型架构下的普遍性。
链接: https://arxiv.org/abs/2607.21804
作者: Run Wang,Chaoyi Zhou,Xi Liu,Yi Zhu,Amir Salarpour,Pedram MohajerAnsari,Zhi-Qi Cheng,Feng Luo,Siyu Huang,Mert D. Pesé
机构: Clemson University (克莱姆森大学); Wayne State University (韦恩州立大学); University of Washington (华盛顿大学)
类目: Cryptography and Security (cs.CR); Computation and Language (cs.CL); Machine Learning (cs.LG)
备注:
Abstract:Lossless acceleration schemes, such as speculative decoding, promise significant inference speedups by relying on dynamic token-level alignment between a draft and a target model. However, this guarantee of semantic equivalence masks a severe operational vulnerability: draft-target alignment can be systematically attacked. In this paper, we introduce ADSD, which, to the best of our knowledge, is the first prompt-suffix attack that collapses verifier acceptance by pushing draft probability mass toward tokens the target is unlikely to accept. ADSD uses Soft-Collapse, a verifier-aligned surrogate derived from the asymmetric speculative acceptance rule, together with a target-preservation objective that discourages obvious task corruption. ADSD successfully generates highly effective adversarial suffixes. On the GSM8K dataset, our attack increases the mean sample time by 62.3% while preserving the task quality. We further show that this vulnerability exists across different domains, speculative decoding strategies, and model architectures.
[NLP-32] Agent ic Evaluation of Copyright Law Compliance ICML2026
【速读】： 该论文旨在解决生成式 AI（Generative AI）代理在执行涉及外部内容检索与再现的商业任务时，如何遵守版权法（copyright law）的问题。当前缺乏有效的评估框架来衡量此类代理在实际应用中对版权法规的合规性。为此，研究提出了 Copyright-Bench 基准测试体系，用于系统评估大语言模型（LLM）代理在真实商业场景（如网站开发、商品设计和商业演示文稿制作）中对公共领域内容与受版权保护内容的选择行为。其核心解决方案在于构建具有现实意义的任务场景，并引入提示变体以模拟用户偏好及时间压力等动态因素，从而揭示模型在面对法律风险时的行为模式。关键发现表明：（1）即使存在合法的公共领域替代资源，代理仍倾向于选择受版权保护的内容；（2）对于开源权重模型，特定用户偏好和时间压力会显著提升侵权行为的发生率。这一结果凸显了现有生成式 AI 代理在版权合规方面的重大缺陷，亟需更严格的法律与技术约束机制。
链接: https://arxiv.org/abs/2607.21799
作者: Zheng Hui,Doni Bloomfield,Noam Kolt
机构: 未知
类目: Computation and Language (cs.CL); Computers and Society (cs.CY)
备注: ICML 2026 Spotlight
Abstract:Large language model (LLM) agents increasingly perform commercial tasks that involve retrieving external content such as images and, where appropriate, reproducing that content. LLM agents should comply with the law, including copyright law. Presently, however, we lack adequate frameworks to assess whether they do so in practice. To that end, we introduce \textbfCopyright-Bench, a benchmark designed to evaluate \textitLLM agents’ compliance with \emphcopyright law. Copyright-Bench is comprised of realistic commercial tasks—website development, merchandise design, and pitch deck production—that involve agents selecting between public-domain content (the use of which is \textitlegal) and copyrighted content (the use of which is \textitinfringing in this setting).The evaluation introduces prompt variations that simulate different user preferences, as well as time this http URL state-of-the-art LLM agents against a human baseline, we find that: (1) agents select copyrighted works despite the availability of public-domain alternatives; and (2) for open-weights models, violation rates increase in response to certain user preferences and simulated time pressure.
[NLP-33] From Seasonality to Semantics: Benchmarking a Hybrid Probabilistic Forecasting System for Roadblocks in Bolivia
【速读】： 该论文旨在解决玻利维亚频繁发生的道路封锁（roadblock）事件缺乏本地化预测系统的问题，此类事件对国民经济造成重大损失，相当于全国生产总值的4%。现有方法难以提前预警，影响物流决策。其解决方案的关键在于构建一种混合概率预测系统，融合时间序列分解模型（Prophet）与自然语言处理（NLP）技术，通过对六年期玻利维亚新闻语料进行向量语义嵌入和零样本分类分析，捕捉社会话语升级的早期信号。实验采用扩展式滚动验证（expanding walk-forward validation），在1,762天内评估七个预测时序（H+1至H+7），结果显示，结合Prophet与NLP的混合配置（C6）显著优于纯统计模型（如SARIMA、LightGBM），在H+1时点达到0.677的AUC-ROC，并使Brier Score降低10.9%（0.220 vs. 0.247），且在所有预测时序上均保持统计显著性（p < 0.02）。研究证明，通过引入新闻文本中的语义信号，可有效识别历史惯性无法捕捉的社会紧张峰值，为关键交通走廊的风险管理提供了可操作的技术工具。
链接: https://arxiv.org/abs/2607.21785
作者: Rodrigo Vargas Sainz,Christian Berón Curti
机构: Universidad Privada de Santa Cruz de la Sierra (私立圣克鲁斯大学)
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Machine Learning (cs.LG)
备注: 21 pages in English + 21 pages in Spanish, 7 figures
Abstract:Roadblocks in Bolivia are a social conflict phenomenon with devastating economic impacts, estimated at losses equivalent to 4% of the national Gross Domestic Product. Despite their recurrence and impact, there is a lack of local predictive systems to anticipate these events for logistical decision-making. This paper presents a hybrid probabilistic forecasting system that integrates time series decomposition (Prophet) with natural language processing (NLP) techniques applied to a six-year corpus of Bolivian news coverage. The methodology employs vector semantic embeddings and zero-shot classification models to capture signals of discursive escalation prior to the materialization of the roadblocks. Using an expanding walk-forward validation scheme applied over 1,762 days and seven forecasting horizons (H+1 to H+7), seven internal configurations and four external benchmarks were compared, including SARIMA and LightGBM. The results demonstrate that the hybrid configuration (Prophet + NLP, C6) consistently outperforms purely statistical models, achieving an AUC-ROC of 0.677 at H+1 and reducing the Brier Score by 10.9% relative to the baseline temporal model (0.220 vs. 0.247), maintaining a statistically significant error reduction across all evaluated horizons ( p 0.02 ). This research validates that the integration of semantic news signals allows for the detection of social tension peaks not captured by historical inertia, providing a technical tool for risk management in critical transport corridors.
[NLP-34] Khondo: A Multimodal Benchmark for Document Packet Splitting of Bangla Forms
【速读】： 该论文旨在解决低资源语言环境下政府文书文档包（document packet）的自动分割难题，尤其聚焦于孟加拉国行政文件中多文档拼接后难以准确还原原始文档结构的问题。其核心挑战在于：在仅依赖页面图像（vision-native）且缺乏显式文本标注的情况下，如何精确识别文档边界并恢复被随机打乱的页面顺序。解决方案的关键在于构建首个面向孟加拉语—英语双语、以视觉为本的基准数据集Khondo，该数据集涵盖14个行政领域、五种不同的文档拼接方式（从顺序到完全打乱），并提供真实标签的文档边界、领域类型与页面顺序信息。通过零样本评估多模态大模型（MLLMs），研究发现模型虽能较好聚类出原始文档片段，但在处理页面顺序打乱的场景时表现显著下降。进一步控制实验表明，页面顺序重建是主要瓶颈，明确的顺序指令虽必要但不足以完全解决该问题，而语言因素（英语优于孟加拉语）亦构成次要但稳定的干扰项。因此，该研究将“页面顺序重建”确立为基于视觉的低资源文档理解中的关键开放问题，并提供了可复现、可控的基准评测体系以推动后续研究进展。
链接: https://arxiv.org/abs/2607.21780
作者: Abu Tyeb Azad,Fahim Ahmed,Ishita Sur Apan,Ezharuddin Jubaer,Sumaiya Karim Katha,Armun Alam,Amin Ahsan Ali,Aman Chadha,Md Mofijul Islam,AKM Mahbubur Rahman
机构: Wichita State University (威奇托州立大学); CCDS, Independent University, Bangladesh (独立大学, 孟加拉国); Amazon GenAI (亚马逊生成式人工智能)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Document packets, multiple documents concatenated into a single file, are common in government and administrative workflows, yet splitting them into their constituent documents is difficult, especially for low-resource languages. We introduce Khondo (Bangla for split/segment), the first benchmark for document packet splitting on Bangladeshi government forms. Unlike prior English and OCR-text-based datasets, Khondo is bilingual (Bangla–English) and vision-native; where models operate directly on page images. It spans five concatenation schemes, from sequential to fully shuffled, across 14 administrative domains, with ground-truth boundaries, domain types, and page order. Zero-shot evaluation of MLLMs shows they cluster pages into their source documents fairly well but struggle in restoring the original page order once shuffled. To isolate what drives this difficulty, we run two controlled analyses, varying the prompt instruction and then the packet language. Both primarily affect ordering rather than clustering: (a) explicit page-order instructions are necessary but insufficient, and (b) English packets are ordered more reliably than Bangla, making page arrangement the dominant challenge and language a secondary but consistent factor. Khondo establishes page-order reconstruction as a key open problem in vision-based, low-resource document understanding, and provides a controlled benchmark for measuring progress toward solving it. Our dataset and code is available at this https URL
[NLP-35] Probing Latent Colombian Identity Inferences in Qwen 2.5-7B with Natural Language Autoencoders
【速读】： 该论文旨在解决大语言模型（Large Language Models, LLMs）在处理非显式提及哥伦比亚身份、社会经济地位或刻板印象相关语境时，是否会在内部表征中隐含这些属性的问题。研究聚焦于模型对西班牙语（哥伦比亚语境）与英语提示的响应，探究其在未明确表达的情况下是否存在潜在的国籍或刻板印象表征。解决方案的关键在于采用自然语言自编码器（Natural Language Autoencoders, NLA）对模型第20层残差流激活值（residual-stream activations）进行解码，通过分析四个位置四分位点上的激活模式，将隐藏的语义信息“口语化”以揭示潜在的隐含表征。该方法实现了从激活层面的可解释性分析向偏见评估的延伸，尤其关注低资源西班牙语变体在模型内部的表示状态，为识别和理解模型中针对非主流语言变体的隐性偏见提供了新的技术路径。
链接: https://arxiv.org/abs/2607.21774
作者: Pablo Santiago Potes Velasco,María del Mar García Matabanchoy,Óscar Julián Pérez Ladino,Jhoan Stevan Mosquera Ortiz,Nicolás Lozano Mazuera,Gilber Alexis Corrales Gallego
机构: Universidad Autónoma de Occidente (奥斯特大学); GobLab-UAI (政府实验室-乌艾)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注:
Abstract:Large language models may infer demographic attributes from subtle linguistic cues even when those attributes are not explicitly stated. This pilot study examines whether Qwen2.5-7B-Instruct internally represents Colombian identity, socioeconomic status, or stereotype-related information when processing Colombian-Spanish and English prompts. We use Natural Language Autoencoders (NLA) to verbalize residual-stream activations from layer 20 across four positional quartiles per prompt. Our dataset contains 30 prompts arranged as 15 matched Spanish-English pairs, spanning explicit Colombian cues, implicit Colombian cues, and neutral controls. We report descriptive rates and qualitative evidence rather than statistically powered effects, focusing on whether latent nationality or stereotype representations appear before they are verbalized in the model output. This work connects activation-level interpretability with bias evaluation for underrepresented Spanish varieties.
[NLP-36] Humanly: A Configurable and Traceable Environment for Human-AI Collaborative Writing
【速读】： 该论文旨在解决在教育与学术评审场景中，评估者仅基于最终文本难以判断其创作过程的问题，即无法区分文本是由人工书写、生成式AI（Generative AI）生成，还是人机协同完成。现有过程追踪工具存在依赖文档历史、活动记录粗粒度以及对写作环境控制能力有限等缺陷。本文提出的解决方案——Humanly平台，将写作过程本身作为可验证的证据，通过用户自定义写作环境并在此环境中进行创作，系统自动记录完整的写作行为及平台内AI辅助信息。其关键创新在于能够将完成的写作会话封装为带有配置感知的异常行为审查功能的“写作证书”（writing certificate），从而实现对写作过程的可信追溯与验证。用户研究表明，Humanly在教师、会议主席和公众读者等多重角色中均具实用性；红队测试进一步验证了其内置的“人类打字检测器”（Humanly Typing Detector）能有效区分人工输入与自动化输入，具备较强的抗伪造能力。
链接: https://arxiv.org/abs/2607.21758
作者: Shenzhe Zhu,Haoqian Zhang,Xu Yang,Jingyu Tang,Yi Nian,Xiaoxue Du,Shu Yang,Alex Pentland,Joachim Baumann,Jiaxin Pei
机构: UT Austin (德克萨斯大学奥斯汀分校); University of Toronto (多伦多大学); Stanford University (斯坦福大学); MIT (麻省理工学院); USC (南加州大学); KAUST (卡塔尔科技大学)
类目: Computation and Language (cs.CL)
备注:
Abstract:Teachers, conference chairs, and public readers all judge writing from limited evidence, seeing only a finished document and not the process that produced it. Final text alone cannot reveal whether a document was produced through human typing, AI generation, or mixed human-AI collaboration. Existing process-tracking tools help, but many are tied to host-document histories, provide coarse activity records, and offer limited control over the writing environment. Humanly is a writing platform that makes the writing process itself the evidence. Users configure writing environments for personal documents or assigned tasks and draft in a workspace that records writing activity and in-platform AI assistance. Humanly can package a completed session into a sealed writing certificate with configuration-aware anomaly behavior review. It can support writing scenarios such as course assignments, peer review, and personal certification. Our user study shows that Humanly is helpful across roles, and a red-teaming study shows that the Humanly Typing Detector distinguishes human hand typing from automated typing.
[NLP-37] Learning What Matters: Supervising Sparse Attention Routing with Causal Evidence Sets
【速读】： 该论文旨在解决生成式AI中稀疏注意力（Sparse Attention）机制在长上下文处理时存在的核心问题：当前基于教师模型注意力模式进行蒸馏的稀疏选择器，其训练假设——注意力分布反映了模型实际依赖的上下文信息——可能不成立。研究发现，在检索任务中，注意力与因果依赖性（causal dependence）常出现显著分歧，导致蒸馏得到的选择器继承了这种偏差。具体表现为：教师模型会关注已学习忽略的过时事实，且在不同训练轮次间注意力分布不稳定；在两步引用任务中，注意力跳过中间推理步骤（因该步骤已在前向传播中完成），致使基于注意力训练的选择器准确率仅为41%，而基于因果证据训练的选择器达到99%并完全匹配教师性能。进一步实验表明，无需人工标注，仅通过掩码冻结教师模型即可恢复有效证据集，并以此训练选择器达到相同高精度。在预训练模型（如Qwen2.5-3B和Gemma-2-9B）中也观察到类似现象：模型虽能正确回答问题，但注意力仍偏向过时信息，而限制其仅访问相关句子后，准确率显著提升。研究结论表明，注意力反映的是模型“注视”的位置，而非其决策所依赖的真实因果关系；在所测试的各种场景中，因果依赖性作为训练目标，其表现至少与注意力相当，甚至更优。因此，解决方案的关键在于：以因果依赖性替代注意力作为稀疏选择器的训练目标，从而实现更可靠、更高效且更符合逻辑的上下文压缩。
链接: https://arxiv.org/abs/2607.21692
作者: Jim Allchin
机构: 未知
类目: Machine Learning (cs.LG); Computation and Language (cs.CL)
备注: 18 pages, 4 figures, 15 tables
Abstract:Sparse attention reduces the cost of long contexts by allowing each query to read only selected parts of the input. These selectors are often trained by distilling the attention patterns of a dense teacher, assuming that attention reveals which context the teacher actually uses. We test that assumption on retrieval tasks where the evidence for each answer is known exactly. By masking parts of the context and measuring whether the answer changes, we find that attention and causal dependence often disagree, and distilled selectors inherit the mismatch. Teachers attend to outdated facts they have learned to ignore, and their attention can vary across training runs even when they rely on the same evidence. In a two-step reference task, attention at the answer skips the intermediate step because it was resolved earlier in the forward pass: a selector trained on attention achieves 41% accuracy, while the same selector trained on causal evidence reaches 99% and matches the teacher. These evidence sets require no annotation: recovered from a frozen teacher by masking alone, they train selectors to the same accuracy. We find the same conflict in pretrained models: Qwen2.5-3B gives more attention to an outdated fact than the current one on 58% of conflicting-fact examples despite answering correctly, while Gemma-2-9B rises from 56% to 99% accuracy when restricted to the two relevant sentences. Attention shows where a model looks, not necessarily what its answer depends on; across the regimes we tested, that dependence matched or outperformed attention as a training target.
[NLP-38] Evaluation design conditions the expert-vs-auto MeSH gap: a controlled comparison of bag-of-words and BiomedBERT on the Cohen benchmark
【速读】： 该论文旨在解决在系统性综述中，如何有效利用医学主题词（MeSH）作为分类器输入特征以提升文献筛选效率的问题，特别是比较由专家标注的MeSH与自动工具生成的MeSH在分类性能上的差异。其核心挑战在于：现有研究未直接对比这两种MeSH来源作为分类器特征的有效性，且缺乏对评估设计（如数据划分方式、交叉验证策略）是否影响结果的系统分析。解决方案的关键在于通过在三个主题（他汀类药物、阿片类药物、注意力缺陷多动障碍）上使用经典药物分类基准（Cohen et al., 2006），分别训练基于词袋模型的逻辑回归分类器（七次重运行）和BiomedBERT（五次种子初始化），并系统考察不同评估设计（包括5折全语料、10折交叉验证、小规模语料匹配等）对分类性能差距的影响。研究发现，在标准5折评估下，专家标注MeSH相较于自动提取MeSH带来+0.096 WSS@95%的性能优势，但当调整语料规模或改用10折交叉验证后，该差距显著缩小至+0.033或+0.021，且置信区间包含零；同时，BiomedBERT在标准评估下仅表现出+0.020的优势，与词袋模型的10折结果处于采样误差范围内。进一步的效能分析表明，对于阿片类药物和注意力缺陷多动障碍主题，因方差较大导致效应量无法被检测到，说明其“无显著差异”的结论实为评估设计限制所致。此外，还观察到表示层面的不对称性：15.1%的他汀类药物样本在附加专家MeSH后超出BiomedBERT的512词元限制，可能因截断而缩小了变压器模型的性能差距，但无法与训练量差异分离。因此，研究结论强调，分类器性能的特征源判断高度依赖于评估设计，合理变更评估方案可能导致结论发生实质性变化。
链接: https://arxiv.org/abs/2607.21685
作者: Samuel M. Okoe-Mensah
机构: 未知
类目: Computation and Language (cs.CL)
备注: 15 pages, 2 figures, 10 tables
Abstract:A systematic review begins with someone reading thousands of abstracts to identify the few that are relevant, and classifiers are used to prioritise that reading. Their inputs are often augmented with Medical Subject Headings (MeSH), assigned either by expert indexers weeks or months after publication or by automatic tools at once. To our knowledge the two have not been compared directly as classifier features, and no previous work has asked whether that comparison’s outcome depends on how the classifier is evaluated. Using the Cohen et al. (2006) drug-class benchmark on three topics, we characterise a bag-of-words logistic regression classifier (seven reruns) and BiomedBERT (five seeds), then examine how the Statins result changes under alternative designs. Under the canonical 5-fold full-corpus design, the bag-of-words expert-vs-auto gap on Statins is +0.096 WSS@95%. Matching the corpus size to the smaller topics (n = 803) reduces it to +0.033 (95% bootstrap CI includes zero), and 10-fold cross-validation at full size to +0.021 (CI narrowly excludes zero). Under canonical evaluation BiomedBERT gives +0.020, within sampling noise of the bag-of-words 10-fold result. A power analysis indicates a Statins-sized effect would not have been detectable at the Opioids or ADHD variance, so those nulls are design-limited rather than informative. A representation asymmetry remains: 15.1% of Statins inputs exceed BiomedBERT’s 512-token limit when expert MeSH terms are appended, so truncation may contribute to the smaller transformer gap, although this cannot be separated from training volume here. In screening pipelines using transformers or 10-fold bag-of-words, the gap on the topics tested is about 0.02 WSS@95%, with CIs spanning zero on at least one bound. More broadly, benchmark conclusions about feature sources can change substantially under reasonable changes to the evaluation design.
[NLP-39] Progress Reward Modeling for Robotic Learning: A Comprehensive Survey
【速读】： 该论文旨在解决机器人学习中因缺乏中间阶段反馈而导致的进展评估难题。在动态环境中，仅依靠任务完成的终端成功信号无法提供行为是否有效推进、停滞或倒退的判断，限制了学习效率。为此，现有研究逐渐转向引入进度奖励（progress reward），以在任务执行过程中提供持续反馈。然而，当前方法在观测输入、目标设定、输出信号形式、监督来源及评估协议等方面缺乏统一框架，导致难以进行公平比较与结果验证。本文提出一个统一的进度奖励建模视角，将其系统性地分解为三个相互关联的步骤：首先从外部接口层面定义进度模型的输入信息与输出信号形式，明确问题边界；其次深入模型内部，分析进度信号构建的方法及其背后的假设与机制；最后考察支撑这些方法的数据集与基准测试，揭示进度监督的获取方式及评估指标的实际含义。这一三重框架实现了对进度模型“是什么、如何构建、如何验证”的完整连接。文章进一步总结了现有方法的核心局限，并展望了未来的研究方向。
链接: https://arxiv.org/abs/2607.21655
作者: Jianshu Zhang,Keliang Wu,Haoran Lu,Anbang Liu,Ce Zhang,Weijie Yin,Chengxuan Qian,Xiyuan Yang,Zhenyu Pan,Guo Ye,Han Liu
机构: 未知
类目: Robotics (cs.RO); Computation and Language (cs.CL)
备注: Project page: this https URL
Abstract:Robotic learning takes place in dynamic environments with large behavior spaces. A terminal success signal only tells the robot whether the task is completed. It does not explain whether the current behavior is making progress, remaining unchanged, or undoing earlier progress. For this reason, recent studies have increasingly explored progress rewards that provide feedback during task execution. However, the current literature lacks a shared framework. Existing methods use different observations, goal specifications, output signals, supervision sources, and evaluation protocols. This makes it difficult to compare them and understand what their results actually validate. In this survey, we provide a unified view of progress reward modeling for robotic learning. We organize the field in three connected steps. We first study the interface of a progress model. This defines the problem from the outside by asking what information the model receives and what form of progress signal it produces. We then move inside the model and study the methods used to construct this signal. This reveals the different assumptions and mechanisms behind progress estimation and reward generation. Finally, we examine the data and benchmarks that support these methods. This shows how progress supervision is obtained and what different evaluations actually measure. Together, these three perspectives connect what a progress model is, how it is built, and how its quality is validated. We further summarize the main limitations of current approaches and discuss future research directions.
[NLP-40] Molt: A Scalable PyTorch-Native Training Framework for Agent ic Reinforcement Learning
【速读】： 该论文旨在解决当前**智能体强化学习（Agentic Reinforcement Learning）**研究中因算法迭代频繁而带来的开发成本过高问题。具体而言，主流框架中每次算法变更（如新估计器、新流水线阶段或新回放策略）均需在训练器、分布式后端及回放接口等多个层级进行修改，导致研究人员在每轮迭代中承受巨大工程负担。其解决方案的关键在于设计并实现一个基于PyTorch原生的轻量级训练框架——Molt，该框架通过保持代码库简洁、可读性强，使研究者能够将整个算法流程完整地掌握于心，同时便于AI编程助手进行全局理解和推理，从而实现从算法逻辑到执行实现的端到端可追溯与可修改性。此外，Molt采用单一异步循环，支持多模态与专家混合（Mixture-of-Experts）策略的联合训练，且确保模型仅基于自身生成的token进行训练，保证了训练过程在token、策略版本和模型语义层面的一致性。实验证明，该框架在采用相同全异步协议下，性能与最先进的基于Megatron的系统统计相当，实现了“轻量化”与“高性能”的统一。
链接: https://arxiv.org/abs/2607.21653
作者: Jian Hu,Huiying Li,Hao Zhang,Binfeng Xu,Yifan Zhang,Shaokun Zhang,Hemil Desai,Michael Demoret,Pavlo Molchanov,Jan Kautz,Yi Dong
机构: NVIDIA(英伟达)
类目: Machine Learning (cs.LG); Computation and Language (cs.CL); Distributed, Parallel, and Cluster Computing (cs.DC)
备注: tech report for Molt
Abstract:Agentic reinforcement learning research is constant algorithm modification, new estimators, new pipeline stages, new rollout schemes, and in mainstream frameworks each change threads through layers of trainer, distributed backend, and rollout glue: the cost lands on the researcher at every iteration. Molt is a PyTorch-native training framework built to keep that cost small: a codebase compact and clean enough for a researcher to hold in their head, and for an AI coding assistant to read and reason about in its entirety, so the algorithm flow can be traced and changed end to end. The agent is an ordinary program, and one asynchronous loop trains multimodal and mixture-of-experts policies while never training on a token it did not generate, consistent in tokens, policy versions, and model semantics. Leanness does not cost performance: under a matched, fully asynchronous protocol, Molt is statistically comparable to a state-of-the-art Megatron-based stack. Molt is open source and provides recipes and containers at this https URL.
[NLP-41] A Consensus-Based Framework for Relative Preference Evaluation of Large Language Models
【速读】： 该论文旨在解决传统大语言模型（LLM）评估体系中因依赖静态数据集和客观评分标准而导致的局限性问题，特别是在存在多个合理回答时，难以有效区分响应在清晰度、完整性和实用性等方面的差异。其核心解决方案是提出一种基于共识的评估框架，通过多模型间的相对偏好判断替代对绝对正确性的衡量。该方法不依赖固定真值，而是让一组多样化的先进大语言模型在盲态条件下对同一提示生成的匿名候选响应进行独立排序与投票，将模型间的一致性聚合为“相对智能指数”（Relative Intelligence Index, RII），以此反映模型输出被其他模型偏好接受的频率。该框架实现了可扩展、模型驱动的比较评估，能够在多种任务领域（如编程、通用知识、安全性、逻辑推理与数学）中揭示稳定的偏好模式，虽不直接对应人类判断，但已有研究表明模型偏好具有与人类评价部分相关的潜力，因而可作为多解场景下响应质量的代理信号。
链接: https://arxiv.org/abs/2607.21632
作者: Mohtashim Khan
机构: 未知
类目: Computation and Language (cs.CL)
备注: 14 pages, 7 figures
Abstract:Traditional benchmarks for LLMs primarily rely on static datasets and objective scoring metrics, which often fail to capture differences in response quality when multiple answers are acceptable. In such settings, correctness alone is insufficient to distinguish between responses that vary in clarity, completeness, and usefulness. This paper introduces a consensus-based evaluation framework that measures relative preference among model-generated responses rather than absolute correctness. Instead of evaluating outputs against a fixed ground truth, we assess how a panel of diverse LLMs ranks anonymized candidate responses to the same prompt. This approach treats aggregate inter-model agreement as a proxy for perceived response quality under blind conditions. We conduct a controlled study using five state-of-the-art LLMs across multiple domains, including programming, general knowledge, safety, logical reasoning, and mathematics. Each model generates responses and independently ranks peer outputs through a structured voting process. Scores are aggregated into a Relative Intelligence Index (RII), representing how frequently a model’s responses are preferred by other models. Our findings reveal consistent preference patterns across domains, with certain models more frequently ranked highly by their peers. However, we emphasize that these results reflect inter-model preference alignment rather than objective correctness or human judgment. This framework provides a scalable, model-driven method for comparative evaluation, offering an alternative perspective on response quality in scenarios where multiple valid answers exist. While not directly aligned with human evaluation, prior work suggests that aggregated model preferences can partially correlate with human judgments, motivating this as a proxy signal. Comments: 14 pages, 7 figures Subjects: Computation and Language (cs.CL) MSC classes: I.2 Cite as: arXiv:2607.21632 [cs.CL] (or arXiv:2607.21632v1 [cs.CL] for this version) https://doi.org/10.48550/arXiv.2607.21632 Focus to learn more arXiv-issued DOI via DataCite Submission history From: Mohtashim Khan [view email] [v1] Sun, 19 Jul 2026 08:59:19 UTC (463 KB)
[NLP-42] Adversarial Style Optimization: Enhancing VLM Jailbreaks by GRPO-based Stylistic Triggers Optimization CVPR2026
【速读】： 该论文旨在解决多模态大语言模型（Multimodal Large Language Models, MLLMs）在安全对齐方面易受越狱攻击（jailbreak attacks）的问题，尤其针对现有基于内容的越狱方法在面对快速演进的MLLMs时表现不稳定、难以有效利用非内容层面漏洞的局限性。其核心发现是：MLLMs在语义理解能力与安全防御能力之间存在显著的风格不一致性（Stylistic Inconsistency）——即模型对视觉风格变化具有鲁棒的语义理解能力，但其安全机制却容易被特定风格触发所绕过。为此，作者提出一种即插即用的增强模块Adversarial Style Optimization (ASO)，通过微调图像编辑模型，在给定对抗性图像上叠加经过优化的风格扰动，从而放大现有视觉越狱攻击的效果。ASO采用由结构化分层奖励函数（Structurally-Tiered Reward Function）指导的组相对策略优化（Group Relative Policy Optimization, GRPO）代理，该奖励函数融合了基于逻辑值（logit-based）信号以检测明确拒绝行为，以及来自强大判别模型的高保真语义评估。大量实验表明，ASO能显著提升当前最优攻击的越狱成功率（Attack Success Rate, ASR），证实了风格偏差作为红队测试MLLMs可扩展攻击向量的有效性。
链接: https://arxiv.org/abs/2607.21619
作者: Bingjun Luo,Jialin Guo,Yue Yao,Xinpeng Ding
机构: Tsinghua University (清华大学); Harbin Engineering University (哈尔滨工程大学); Shandong University (山东大学); Xidian University (西安电子科技大学)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: Accepted by CVPR 2026 (Oral)
Abstract:Multimodal Large Language Models (MLLMs) have achieved impressive performance, but their safety alignment remains vulnerable to jailbreak attacks. Existing content-based jailbreaks are often inconsistent and show unsatisfying performance against the rapidly evolving MLLMs, failing to exploit non-content-based vulnerabilities. Unlike previous research, we empirically find that MLLMs exhibit a Stylistic Inconsistency between their comprehension ability and safety ability: MLLMs can robustly understand content regardless of visual style, yet their defense mechanisms can be easily bypassed by specific stylistic triggers. Based on this finding, we propose Adversarial Style Optimization (ASO), a plug-and-play enhancement module to amplify existing visual jailbreaks. ASO fine-tunes an image-editing model to superimpose an optimized stylistic modification onto a given adversarial image, using a Group Relative Policy Optimization (GRPO) agent guided by a Structurally-Tiered Reward Function that combines a logit-based signal for detecting explicit refusals with a high-fidelity semantic evaluation from a powerful judge model. Extensive experiments show that ASO significantly enhances the ASR of SOTA attacks, demonstrating that stylistic biases are a scalable vector for red-teaming MLLMs. Our code is available at this https URL.
[NLP-43] Do VLMs Read or Rewrite? On Transcription Faithfulness in Vision-Language Models
【速读】： 该论文旨在解决生成式视觉语言模型（Vision Language Models, VLMs）在文档理解任务中对文本不完美输入的处理缺陷问题，即当输入文本存在噪声或错误时，这些模型倾向于将其“重写”为更符合语义合理性但可能偏离原始内容的形式，而这一行为在传统清洁文本评测基准中难以被发现。其核心解决方案是提出一个名为FaithC4的多语言扰动基准，涵盖1,455页单页文档（英语、中文、韩语），包含三种扰动类型：打乱（scramble）、随机替换（random substitution）和视觉相似替换（visually similar substitution）。通过该基准评估15种系统（包括通用VLM、OCR专用VLM及传统OCR流水线），揭示三类系统在扰动下的表现差异：通用VLM的词错误率（WER）恶化最高达4.5点，OCR专用VLM为0.2–2点，传统OCR则低于0.6点（英文）。进一步对Qwen3-VL-4B进行逐层探查发现，重写行为仅在扰动词的最终前馈网络（FFN）表示与原始编码保持接近时触发；一旦表示充分发散，则模型转为忠实转录。此外，词长显著影响重写概率：4–6字符短词重写率可达10%，而超过8字符后重写率骤降至0%。该研究的关键在于揭示了模型重写行为的内在机制，并提出了可量化评估模型忠实性的新基准。
链接: https://arxiv.org/abs/2607.21617
作者: Gwang Gook Lee,Kenan Emir Ak,Jay Mohta,Yan Xu,Dimitrios Dimitriadis
机构: Amazon.com
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注: 15 pages, 6 figures
Abstract:Vision Language Models (VLMs) are increasingly used in place of traditional OCR pipelines for document understanding. In this paper, we show they do not always act as faithful transcribers: when text is imperfect, they often tend to rewrite it into a more plausible form - a behavior that clean-text OCR benchmarks cannot detect. We introduce FaithC4, a multilingual perturbation benchmark of 1,455 single-page documents (English, Chinese, Korean) with three perturbation families: scramble, random substitution, and visually similar substitution. We use the benchmark to evaluate 15 systems spanning general-purpose VLMs, OCR-specialized VLMs, and traditional OCR pipelines. These three categories differ in WER degradation under perturbation: general-purpose VLMs degrade by up to 4.5 points, OCR-specialized VLMs by 0.2-2 points, and traditional OCR by less than 0.6 points on English. Probing Qwen3-VL-4B layer-by-layer, we identify a consistent pattern: rewriting fires only when a perturbed word’s final layer FFN representation stays close to the original encoding; when the representation diverges sufficiently, the model transcribes faithfully. Word length affects rewriting rate: short words (4-6 characters) are rewritten up to 10% of the time, with a sharp cutoff at 8 characters above which rewriting drops to 0%.
[NLP-44] he Hard Decision Layer: Evidence for Committed Inference in Transformers
【速读】： 该论文旨在解决生成式语言模型在多项选择题问答任务中决策过程的可解释性问题，具体聚焦于模型在推理过程中何时以及如何对答案选项进行最终判定。其核心发现是识别出一个名为“硬决策层”（Hard Decision Layer, HDL）的自然架构特性——在该层，模型对候选答案的排序会突然且显著地趋于稳定，标志着模型已做出关键预测。解决方案的关键在于揭示这一现象不依赖于人为设计的路由策略，而是由模型架构本身决定，并在多个主流大语言模型（Qwen、Llama、Granite、Mistral）及多个基准数据集上表现出一致性和鲁棒性；此外，研究还发现HDL在微调后保持不变，且在该层附近模型准确率出现显著跃升（最高达+0.61，如Qwen在CommonsenseQA上的表现），之后性能趋于平稳。系统性消融实验进一步验证了该现象与标签格式和问题复杂度无关，具有根本性的架构基础。这些发现为理解Transformer模型的推理机制提供了新的机理视角，并为高效推理优化与模型引导（model steering）提供了潜在路径。
链接: https://arxiv.org/abs/2607.21613
作者: Ashwath Vaithinathan Aravindan,Mayank Kejriwal
机构: University of Southern California (南加州大学); Information Sciences Institute (信息科学研究所)
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注:
Abstract:We investigate where and how transformer-based language models commit to predictions in multiple-choice question answering. We identify the Hard Decision Layer (HDL), a natural architectural property where answer option rankings stabilize abruptly during inference. Empirical validation across four language models (Qwen, Llama, Granite, Mistral) and four benchmark datasets demonstrates consistent HDL emergence without learned routing policies. We also show that the HDL is invariant to fine-tuning. Our results reveal striking accuracy improvements at the HDL: up to +0.61 (Qwen on CommonsenseQA), after which performance stabilizes. Systematic ablations on label formats and problem complexity confirm the phenomenon is fundamental to model architecture. These findings offer mechanistic insights into transformer inference and suggest opportunities for efficient reasoning and model steering. All code and results required to reproduce this work are available in this https URL
[NLP-45] From Obligation to Specification: A Survey on Validating EU AI Act Requirements in RE
【速读】： 该论文旨在解决在欧盟《人工智能法案》（EU AI Act）生效背景下，组织在需求工程（Requirements Engineering, RE）中如何将法规所要求的透明性、风险管理与可追溯性等义务转化为可测试、可审计的需求及可验证证据的问题。当前许多组织缺乏系统化流程来实现这一转化，导致合规准备不足。其解决方案的关键在于利用基于大语言模型（LLM）的代理型闭环验证工具（agentic closed-loop validation tools），通过自动化支持法规义务到具体需求的映射、覆盖率评估以及证据组织，从而提升需求工程的合规能力。研究发现，尽管参与者普遍认可此类工具的潜力，但对完全自动化存在显著担忧，强调必须引入人工监督与安全机制。基于此，论文提出了面向欧盟《人工智能法案》合规的闭环需求工程方法的最低要求。
链接: https://arxiv.org/abs/2607.21608
作者: T.Y. Emmy Lai,Sven Giesselbach,Matthias Koch,Héctor Allende-Cid
机构: University of Chile (智利大学); German Research Center for Artificial Intelligence (德国人工智能研究中心)
类目: oftware Engineering (cs.SE); Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Computers and Society (cs.CY)
备注:
Abstract:With the EU AI Act entering into force, organizations developing or operating AI systems face new obligations on transparency, risk management, and traceability. For Requirements Engineering (RE), these obligations must be translated into testable, auditable requirements and verifiable evidence. However, many organizations currently lack systematic processes to achieve this. We hypothesize that LLM-based agentic validation tools can support this translation, thereby helping to close this gap. We present a mixed-method exploratory study with expert interviews (N=10) and an online survey (N=15) to assess organizational preparedness for EU AI Act-oriented RE and perceptions of LLM-based, agentic closed-loop validation tools, with participants spanning RE, data science, development, and compliance roles. Our results show that, although the EU AI Act is viewed as highly relevant, structured mechanisms to capture regulatory obligations, propagate updates into projects, and maintain lifecycle-wide traceability and evidence are often missing. Participants see LLM-based tools as promising for mapping obligations to requirements, assessing coverage, and organizing evidence, but express strong concerns about full automation and stress the need for safeguards. Based on these findings, we outline minimum requirements for an EU AI Act-ready closed-loop approach.
信息检索
[IR-0] PinEqualizer: Full Funnel Content Exploration and Debiasing System at Pinterest KDD2026
链接: https://arxiv.org/abs/2607.22518
作者: Olafur Gudmundsson,Bo Zhao,Huayi Liao,Anna Kiyantseva,Sai Xiao,Heath Vinicombe,Mostafa Keikha,Luke DeLuccia,Zihao Chen,Junpeng Hou,Weijie Jiang,Bhawna Juneja,Andreanne Lemay,Wei-Ting Lin,Keyvan Moghadam,Jiaxing Qu,Zhiqing Rao,Zhihua Zhang
类目: Information Retrieval (cs.IR); Machine Learning (cs.LG)
备注: 10 pages, 2 figures. Accepted at KDD 2026
Abstract:In this paper, we propose a new solution for addressing the content cold-start problem in industry-scale search and recommender systems. Compared to prior approaches, we have made the following new contributions: 1) our solution spans the entire multi-stage funnel and generalizes well for both search and recommendation surfaces, 2) our solution reduces bias favoring existing content, allowing more accurate model prediction across content types and reducing short-term tradeoffs associated with high volumes of explicit content exploration, 3) our solution is evaluated with a scalable measurement framework that enables fast short-term experimentation while validating long-term impact. We have iteratively built and successfully deployed this new system at Pinterest in the past two years and observed significant improvements in fresh content exploration, overall user engagement, and content ecosystem health.
[IR-1] Legal Nugget Extraction for Granular Retrieval over Long Jurisprudential Texts
链接: https://arxiv.org/abs/2607.22479
作者: Lucas Pereira,Erick Brito,Roberto Lotufo,Jayr Pereira
类目: Information Retrieval (cs.IR)
备注:
Abstract:Legal retrieval over jurisprudential collections is challenging because court decisions are long, heterogeneous documents whose relevant legal thesis may occupy only a small portion of the text. This paper asks whether legal nuggets, defined as short and self-contained legal theses extracted from source documents, can improve dense retrieval over Brazilian legal collections. We propose a pipeline that extracts nuggets from each document, indexes them with embeddings, retrieves nugget-level evidence, and aggregates the retrieved nuggets back to document-level rankings. We evaluate this approach on four Portuguese legal retrieval benchmarks from the JUA ecosystem, reporting NDCG@10, MAP@10, and MRR@10. Nugget retrieval substantially improves the two jurisprudential datasets: on JUA-Juris, NDCG@10 increases from 0.10265 to 0.20461, and on JurisTCU from 0.20898 to 0.32696. However, it underperforms full-document retrieval on NormasTCU and BR-TaxQA, and an embedding-model ablation shows that strong domain-adapted retrievers can remain better in the full-document setting. The results demonstrate that legal nuggets can be useful for jurisprudence search, especially when queries are formulated as legal theses, but they may not transfer equally well to other legal retrieval scenarios.
[IR-2] Reflector: Arrangement-Aware Harmonic Retrieval for Sample-Based Composition
链接: https://arxiv.org/abs/2607.22413
作者: Austin Rockman
类目: ound (cs.SD); Information Retrieval (cs.IR); Machine Learning (cs.LG)
备注: 15 pages, 7 figures, 1 table, code, application, and other resources at this https URL
Abstract:Sample retrieval tools can help composers find harmonically compatible material, but querying from a fixed reference sample becomes less informative as arrangements evolve and the harmonic context shifts with each musical decision. We present Reflector, an interactive audio workstation that tracks harmonic combinations as they accumulate on the composer’s timeline and adapts retrieval as the arrangement develops. The system is organized around a fixed interval-class oracle: a hand-designed table of weights that scores how pitch-class content combines between sources. An encoder trained entirely on synthetic audio learns to approximate the oracle in a 128-dimensional embedding space, where dot products stand in for compatibility scores at interactive speed. As the composer arranges material on a multi-track timeline, a sweep-line analysis discovers co-sounding regions, computes oracle-weighted centroids, and retrieves against the composite harmonic identity of the session as it evolves. Session centroids projected into a navigable 3-D space reveal structural harmonic relations across the composer’s body of work. This paper is a systems account: we give the design rationale for each architectural decision, characterize Reflector’s behavior through intrinsic measurements on a working sample library, and describe the implementation. The characterization yields a central finding: the learned embedding preserves the kernel’s pairwise judgments while covering the whole library, something the kernel cannot do when used directly as a retrieval rule, because the embedding’s normalized geometry cannot express the degenerate solutions that direct scoring favors. The entire pipeline runs locally with no copyrighted training data. Reflector is free, and the training pipeline is open source.
[IR-3] he Prompt Is Not the Query: How Request State Evolves Across Multi-Turn AI Conversations
链接: https://arxiv.org/abs/2607.22392
作者: Benjamin Tannenbaum
类目: Information Retrieval (cs.IR)
备注:
Abstract:AI-search evaluation commonly treats a prompt as a stable query that can be counted, classified, and replayed in isolation. A conversation makes that unit of analysis questionable: each user turn can add a constraint, revise an assumption, request evidence, or refer to alternatives established earlier. We replace latent “intent” with an observable construct, conversation-conditioned request state, and measure how that state is distributed across user turns. The analysis reuses frozen rules and the governed cohort of a preceding conversation study: 670 English commercial multi-turn conversations in a discovery-replication design and 7,463 public PRISM conversations from 1,389 participants. In the commercial corpus, the final prompt contains a median 35.6% of the session’s unique user-side content vocabulary; in PRISM, the median is 36.4%. The final prompt contains at most half of that vocabulary in 68.4% and 74.3% of conversations, respectively. More importantly, transparent rules detect at least one request-state dimension in history but not in the final prompt in 50.3% of commercial conversations and 44.8% of PRISM conversations. Among dimension-bearing conversations, the final prompt reproduces the full observed dimension set in only 26.1% and 26.2%. At the same time, the final prompt adds a previously unseen dimension in 17.9% and 19.3%, showing that the endpoint is neither a summary nor merely a reference: it is often another state update. Length-matched nulls show that low lexical coverage is largely a consequence of turn length, so vocabulary results are interpreted as information availability, not semantic drift. The categorical results support session-level measurement for AI search. They do not estimate the causal effect of history on model answers.
[IR-4] Bringing GRACE to Recommendation: Fine-Tuning for Sustainable and Accurate Personalization
链接: https://arxiv.org/abs/2607.22341
作者: Yibowen Zhao,Yinan Zhang,Ning Liu,Lizhen Cui,Chunyan Miao
类目: Information Retrieval (cs.IR)
备注:
Abstract:Growing concern about environmental sustainability (e.g., reducing carbon emissions and resource use) and public health has motivated ``green’’ recommender systems that steer users toward more eco-friendly and healthier choices. However, many existing green recommendation approaches require training new models from scratch, incurring substantial computational and energy costs. Reranking-based methods, meanwhile, introduce an additional sorting stage at inference, increasing latency and computational cost. In this work, we propose GRACE (Green Recommendation via Adaptive Conflict-rEsolution), a fine-tuning framework that integrates item-level sustainability signals (e.g., eco-scores or health indices) into pretrained recommendation models. Since these green values are usually discrete and non-differentiable, existing methods often rely on pairwise comparisons to promote greener items. GRACE instead introduces a differentiable approximation that enables direct optimization of the green criterion. To balance sustainability and personalization quality, GRACE further employs a gradient projection mechanism to mitigate conflicts between the green objective and the accuracy objective during fine-tuning. Experiments on real-world datasets demonstrate that GRACE improves sustainability-oriented recommendation outcomes while generally preserving recommendation accuracy through a controllable preference-anchored update mechanism.
[IR-5] LAMAR: An Open Language-Aware Multilingual Alignment Reranker
链接: https://arxiv.org/abs/2607.22042
作者: Seongtae Hong,Youngjoon Jang,Jungseob Lee,Seungyoon Lee,Heuiseok Lim
类目: Information Retrieval (cs.IR)
备注: preprint
Abstract:In multilingual retrieval augmented generation, a retriever can retrieve relevant documents written in multiple languages, which are subsequently reranked before answer generation. However, it remains unclear whether existing multilingual rerankers consider document language when ordering semantically relevant candidates. Our analysis shows that these rerankers do not consistently prioritize documents written in the same language as the query when semantically equivalent documents are available across languages, even though document language can affect answer generation. We release LAMAR, a language aware multilingual cross encoder trained to account for both semantic relevance and language coherence. LAMAR first uses English anchored relevance distillation to establish consistent relevance scoring across multilingual inputs and then applies preference alignment for language coherence to encourage documents written in the same language as the query to receive higher rankings while retaining semantic relevance. In a controlled experiment designed to assess language coherence, LAMAR achieves the best performance overall and across all languages examined individually. LAMAR also remains competitive on established multilingual reranking benchmarks. In practical retrieval settings, LAMAR achieves the best results across all reported metrics when reranking candidates retrieved in the first stage. These results demonstrate that LAMAR accounts for language coherence while achieving strong performance on general multilingual reranking benchmarks.
[IR-6] SIREN (Luring LLM s onto the Rocks): PAIR-Driven Preference Manipulation in Web-RAG Recommenders
链接: https://arxiv.org/abs/2607.21951
作者: Evan Caville,Siamak Layeghy,Billy Sung,Sara Dolnicar,Marius Portmann
类目: Information Retrieval (cs.IR); Cryptography and Security (cs.CR)
备注:
Abstract:This paper investigates the adversarial manipulation of the ranked recommendations produced by web-augmented large language models (LLMs). When an LLM answers a recommendation query by retrieving and reading live webpages, it acts as a recommender, and each retrieved page becomes a potential attack surface. Prior work has examined fabricated products, retrieval poisoning, and rank promotion. However, these studies do not compare how different edits to an already retrieved page change the model’s final ranking while the surrounding source set remains unchanged. To address this gap, we propose SIREN, an automated attacker–judge method that adapts the PAIR jailbreaking loop to competitive rank manipulation, with the goal of moving a chosen entity to rank~1 in an LLM-generated recommendation. SIREN retrieves and captures webpages using Anthropic’s web tools, then iteratively edits a retrieved source using an interpretable taxonomy of 23 content-poisoning techniques. The custom-RAG replay platform keeps the same sources in the same order, so changes in the model’s ranking can be linked to changes in the supplied content rather than to differences in retrieval. Across two production Claude models, SIREN reaches rank~1 in 62 of 124 technique trials nested within eight query–model contexts. The payloads that reached rank~1 were then tested in fresh sessions, where they reproduced the result with a mean success rate of 0.805. Across the evaluated settings, declarative ranking claims and seeded lists were generally more effective than directive-form injections, although the strength of this difference depended on the target model. To the best of our knowledge, this is among the first controlled studies of competitive rank manipulation in production LLMs where the supplied source context is kept fixed.
[IR-7] StARS: Socially Appropriate Robot Actions via a Recommender System-Driven Approach IROS2026
链接: https://arxiv.org/abs/2607.21802
作者: Erencem Ozbey,Fethiye Irmak Dogan,Jin Huang,Hatice Gunes
类目: Robotics (cs.RO); Information Retrieval (cs.IR)
备注: IROS 2026
Abstract:Social appropriateness in human-robot interaction (HRI) is not universal: different people can judge the same robot action differently in the same situation. To capture this inter-subject variability, we reformulate socially appropriate action generation as a preference modelling problem inspired by recommender systems, treating annotators as users, contexts/scenes as items, and appropriateness scores over a set of candidate robot actions as targets. We propose StARS, a novel model-agnostic framework that integrates collaborative filtering with learnable scene representations to generate user-specific appropriateness scores over candidate robot actions. StARS is model-agnostic: it can be integrated with various scene encoders and backbones, enabling personalisation without redesigning the underlying model. We evaluate StARS on two socially aware robotics datasets, MannersDB+ and SocNav1, and analyse robustness under sparse preference feedback. Across datasets and backbones, StARS consistently improves performance and agreement with annotators, supporting personalised action selection aligned with user norms. Our code is publicly available at this https URL.
人机交互
[HC-0] Gridnberg: A Topography-Aware Pedestrian Routing Dataset for New York City
链接: https://arxiv.org/abs/2607.22523
作者: Ariel Noyman
类目: Human-Computer Interaction (cs.HC); Computers and Society (cs.CY)
备注: 5 pages, 4 figures
Abstract:Cities are rarely flat, yet urban network analysis usually represents streets as planar graphs. This simplification affects modeled impedance, route choice, and the interpretation of accessibility, particularly where alternative paths differ in grade. This paper introduces Gridnberg (‘grid-n-berg’, grid and mountain), a topography-aware pedestrian routing dataset for New York City. The dataset enriches the NYCWalks network with vertex-level elevations derived from the New York City Planimetric Database. For each pedestrian-network geometry vertex, the workflow averages selected elevation observations within a 50 m radius, retains segments with complete vertex support, and uses direction-specific cumulative ascent and descent to calculate three routing costs: horizontal distance, a comfort-oriented slope score, and an accessibility-sensitive slope score. The release retains 313184 of 315577 source segments (99.24%). Gridnberg supports reproducible terrain-aware analysis, transparent scenario comparison, and improved pedestrian-network representations in New York and other cities.
[HC-1] Beyond Perspectives: A Trio-Ethnography of Interpretation Evolution in LLM -Supported Programming Education
链接: https://arxiv.org/abs/2607.22463
作者: Jennie Ren,Jordan H. McDowell,Kyrie Zhixuan Zhou
类目: Human-Computer Interaction (cs.HC); Artificial Intelligence (cs.AI)
备注:
Abstract:Generative AI is reshaping programming education, yet educators often infer students’ AI-supported learning from classroom observations alone. This experience report presents a trio-ethnography involving two computing educators with different teaching philosophies and one undergraduate computer science student to examine how these interpretations evolve through dialogue. Across three conversations, the educators reflected on students’ AI use, discussed changes to programming pedagogy, and revisited their assumptions after engaging with the student’s lived experiences. Rather than simply confirming or contradicting the educators’ perspectives, the student’s narratives revealed learning processes that were largely invisible in the classroom, prompting both educators to reconsider assumptions about AI use, assessment, transparency, and programming instruction. We argue that trio-ethnography offers a valuable reflective approach for helping computing educators move beyond observable student behaviors toward a richer understanding of AI-supported learning and for informing instructional adaptation in the era of generative AI.
[HC-2] Unboxing Diffusion Models for the Arts: Interactive Model Bending and Practice-Based Explainability
链接: https://arxiv.org/abs/2607.22428
作者: Ahmed M. Abuzuraiq,Philippe Pasquier
类目: Human-Computer Interaction (cs.HC); Artificial Intelligence (cs.AI); Machine Learning (cs.LG); Multimedia (cs.MM)
备注: Under review for “Explainable AI for the Arts” (N. Bryan-Kinns, Ed.), Springer
Abstract:Explainable AI (XAI) in creative practice can be less about technocentric explanation and more about enabling artists to inspect modify and debug models as part of making Yet largescale texttoimage diffusion systems are typically presented as opaque endtoend tools limiting this kind of material engagement We argue that even large models can function as creative materials when their internal structure is made visible and manipulable To support this we propose a handson approach to explainability centred on experimentation and intervention We instantiate this approach with a model bending and an interactive (inspection) interface integrated into ComfyUIs nodebased workflow including interactive layer selection and intervention controls Through qualitative and quantitative analysis of bending interventions in Stable Diffusion 15 we show how manipulating specific components of a diffusion pipeline produces relatively consistent families of visual effects allowing artists to build practical layerlevel intuition about how different parts of the model shape generated images
[HC-3] Universal BCI Personalization: One API for Frozen EEG Trunks and Foundation Models
链接: https://arxiv.org/abs/2607.22397
作者: Sergey Musienko
类目: Human-Computer Interaction (cs.HC); Machine Learning (cs.LG); Neurons and Cognition (q-bio.NC)
备注: 20 pages, 12 figures. Companion paper on the control layer forthcoming
Abstract:Frozen EEG encoders proliferate; per-model fine-tune defaults do not scale. We present Nimbus Personalizer: one contract encode to Bayesian head to BrainState (optional affine mid-tier) that sits on heterogeneous frozen trunks without a new personalization stack per architecture. Thesis (systems): the contribution is the trunk-agnostic API - not LDA-on-embeddings as an ML novelty - so OEMs integrate once and swap trunks. Evidence: the same surface runs on five classical trunks EEGNet, Shallow, Deep, Conformer, ATCNet x four MI datasets (18 cells) and on a foundation encoder (REVE) under the same Personalizer. Where embedding capacity exists, the head is a cheap default mid-point versus warm-start fine-tune or PEFT, costing orders of magnitude less adaptation wall time while recovering much of the fine-tune accuracy gain; calibration-only-when-clean holds in 12/18 cells. Head gains are supporting evidence that the API is useful where capacity exists. Subject-level confidence intervals identify the clearest dataset and span zero elsewhere. All results are exploratory (subject-level bootstrap, no confirmatory tests); the decision logic for when to escalate adaptation is addressed in our companion work on the control layer.
[HC-4] Kutti AI: A Voice-First Offline-Capable Learning Companion with Real-Time Struggle Detection for Visually-Impaired Children ATC
链接: https://arxiv.org/abs/2607.22377
作者: Kadharmoideen Fadurudeen
类目: Human-Computer Interaction (cs.HC); Computers and Society (cs.CY); Sound (cs.SD)
备注: 8 pages. Voice-first, offline-capable learning companion for visually-impaired children with multi-signal struggle detection and cross-language answer matching. Prototype built at the Half Baked hackathon (English and Tamil)
Abstract:Most educational technology for children is built around visual interfaces, which excludes the many children worldwide who live with visual impairment – an estimated 1.4 million children are blind and many more have low vision. We present Kutti AI, a voice-first learning companion designed so that audio is the primary and sufficient interface: children learn curriculum concepts through spoken conversation, respond by speaking, and receive spoken feedback, with no reliance on visual elements. The system contributes three practical mechanisms for accessible, adaptive learning on commodity mobile hardware: (1) a multi-signal struggle-detection engine that combines response-latency analysis, wrong-attempt tracking, and keyword-based hesitation detection to decide, in real time, when to offer hints or simplify a question; (2) a multi-layered cross-language answer-matching pipeline that combines language-aware translation/transliteration, Levenshtein-based fuzzy matching, and text normalization so that children are not penalized for code-switching or pronunciation variation; and (3) an offline-first speech pipeline using an on-device automatic speech recognition (ASR) model, enabling use in low-connectivity settings common in underserved communities. We describe the architecture, the interaction flow, and the design decisions that prioritize accessibility, and we report qualitative observations from a hackathon prototype supporting English and Tamil. We discuss lessons learned and outline a path toward formal evaluation with target users. Kutti AI illustrates how a small, carefully-engineered voice-first system can lower both accessibility and financial barriers to early education.
[HC-5] An AI-Driven Virtual Patient for Breaking Bad News: An Expert Formative Study on Facial Expression Intensity
链接: https://arxiv.org/abs/2607.22118
作者: Steffen Hauck,Theresa Schell,Sophie Jörg,Jens Grubert
类目: Human-Computer Interaction (cs.HC)
备注:
Abstract:Interactive virtual patients driven by large language models (LLMs) offer scalable solutions for medical communication training, such as breaking bad news. However, designing their emotional expressiveness remains a challenge. This paper presents an AI-driven virtual patient framework combining LLM dialogue with real-time facial animation in virtual reality (VR). We conducted an exploratory, formative evaluation with seven medical experts to gather early feedback and elicit design requirements. The evaluation focused on how variations in facial expression intensity affect perceived realism and the virtual patient’s emotion intelligibility. While descriptive quantitative ratings remained baseline across conditions, qualitative interviews provided deep insights into how experts perceive virtual emotional cues. The findings suggest that experts evaluate emotional realism holistically through multiple verbal and non-verbal channels; isolated facial adjustments are easily overshadowed by dialogue nuances and vocal prosody. Based on these insights, we present a roadmap for future medical training simulators, highlighting the need for synchronized, multi-modal pipelines incorporating body gestures, conversational pauses, and vocal dynamics.
[HC-6] he machine can say it but cannot hear it. Designed affective patterns and the expressive-sensing asymmetry in human-machine communication
链接: https://arxiv.org/abs/2607.22104
作者: Jan K. Argasinski
类目: Human-Computer Interaction (cs.HC)
备注:
Abstract:Affect-adaptive systems increasingly act as communicators that sense a user’s emotion and respond with events meant to change it, closing an affective loop. This vision assumes both that a machine’s affective messages are received and that the bodily channel it monitors carries an intelligible reply-assumptions rarely tested together. In a within-subjects virtual-reality study (N = 20), an autonomous system delivered six empirically derived affective patterns-scripted emotional events distilled from 104 practitioners’ (first responders’) critical incidents-while we recorded the human reply across felt emotion, felt arousal, and autonomic (electrodermal and cardiac) activity. Acting only as an author of designed messages, the machine reliably evoked strong, differentiated emotions: valence fell sharply for every pattern (|dz| = 1.1-1.7), and the patterns produced distinguishable, individually classifiable signatures of anger, fear, and sadness, functioning as a vocabulary of machine-to-human affective messages. Yet the channel the machine would read stayed largely silent. Self-reported arousal did not change, with Bayesian and equivalence evidence for no effect; sympathetic and cardiac arousal moved only for the most perceptually salient events, not for the messages experienced as most powerful; and within individuals, felt valence was decoupled from both bodily registers. This valence-arousal dissociation reveals an expressive-sensing asymmetry: machine agency extends to expression but not perception, and closed-loop designs regulate on a variable their messages neither reliably move nor are legible in. We draw out implications for theories of machine agency and for affect-adaptive design.
[HC-7] What Clinicians Need: Designing Developing and Evaluating an AI-Based Decision Support System for Autism Assessment
链接: https://arxiv.org/abs/2607.22005
作者: Ulrike Schäfer,William Saakyan,Matthias Norden,Fabrizio Kuruc,Peter Sörries,Isabel Dziobek,Claudia Müller-Birn,Hanna Drimalla
类目: Human-Computer Interaction (cs.HC)
备注: 29 pages, 9 figures
Abstract:AI methods promise to support autism spectrum condition (ASC) diagnostics in adults, a complex and time-consuming process, that is characterized by a shortage of specialized clinicians. To date, clinicians’ needs and their interaction with such AI-based support remain underexplored. Our work aims to develop and evaluate an AI-based clinical decision support system (CDSS) for ASC assessment, and to investigate how it impacts clinicians’ decision-making. By interviewing clinicians of varying experience levels, we identified five challenges and derived design strategies. Based on that, we developed SIT-CARE, a CDSS, which provides AI-based recommendations and data visualizations of clinically relevant nonverbal behavior. Through an evaluation study with newly recruited clinicians, we found that SIT-CARE led to different decision paths in regard to the ASC assessment, which are reflected in clinicians’ mental models and decision changes. Overall, SIT-CARE demonstrated potential in improving initial diagnostic assessments, supporting in-depth diagnosis and empowering less experienced clinicians.
[HC-8] VisionPulse: A Virtual Reality System Enabling Accessible Discovery and Navigation for Blind and Low Vision Users
链接: https://arxiv.org/abs/2607.21944
作者: Samuel Martin,Pooyan Fazli,Hasti Seifi
类目: Human-Computer Interaction (cs.HC)
备注: accepted to ACM ASSETS 2026
Abstract:Free exploration is an important aspect of many engaging virtual reality (VR) experiences, yet remains largely inaccessible to blind and low vision (BLV) users due to its reliance on visual feedback. Existing approaches support BLV navigation through prebuilt menus of environment and audio beacons, but offer limited support for free-form discovery. We present VisionPulse, an accessible VR system that enables BLV users to explore virtual environments through natural head and hand movements, combined with auditory, haptic, and text-to-speech feedback. VisionPulse introduces a discovery-driven approach that allows users to progressively uncover regions and objects, alongside navigation support through waypoint guidance and object localization via responsive audio and orientation-based haptics. A study with 12 BLV participants showed a strong preference for VisionPulse’s discovery-based exploration and multimodal feedback, without negatively impacting task performance or perceived workload. Our findings underscore the importance of accessible, free-form VR experiences, and contribute insights for inclusive VR design.
[HC-9] LatentFlow: Visual Analytics for Latent Space Analysis in Molecular Graph Neural Networks
链接: https://arxiv.org/abs/2607.21941
作者: Shiyi Liu,Jiaqing Chen,Nicholas Hadler,Rostyslav Hnatyshyn,Michael W. Mahoney,Talita Perciano,John F. Hartwig,Gunther H. Weber,Ross Maciejewski
类目: Machine Learning (cs.LG); Human-Computer Interaction (cs.HC)
备注:
Abstract:Chemists and materials scientists increasingly use machine learning models, such as graph neural networks (GNNs), to predict properties of molecules and the outcomes of their reactions. Beyond predictive performance, understanding how these models organize chemical information internally in their latent spaces, i.e., the embeddings of the molecules, is critical. Analyzing latent spaces helps diagnose model behavior and assess whether the learned embeddings are organized in ways that reflect meaningful chemical relationships. Unfortunately, existing methods provide limited support for analyzing latent spaces across layers and across different model states (e.g., training epochs, model configurations, and input data), making it difficult to understand how these latent spaces evolve throughout a model or relate to chemical concepts. We present LatentFlow, a visual analytics system developed in collaboration with a domain expert for analyzing latent spaces in molecular GNNs. LatentFlow groups embeddings into clusters and supports exploration of latent spaces by tracking how these clusters change across layers and model states using a modified Sankey diagram. To support interpretation, LatentFlow links these clusters to representative molecules and their shared substructures, and it allows scientists to introduce their own domain knowledge and compare it with the patterns found in the latent spaces. We evaluate LatentFlow through two case studies. The results show that LatentFlow helps scientists understand how latent spaces evolve, identify meaningful molecular patterns, and better interpret model behavior.
[HC-10] owards Reducing Foreign Language Anxiety Using Level-Appropriate Embodied Conversational Agents
链接: https://arxiv.org/abs/2607.21887
作者: Krishan Rajaratnam,Wenbin Gan,Yuan Sun
类目: Human-Computer Interaction (cs.HC); Computation and Language (cs.CL)
备注: 8 pages, 6 figures, published in the proceedings of EDULEARN26
Abstract:Foreign language anxiety (FLA) can be a major barrier to second language acquisition (SLA), especially in conversational contexts. With the proliferation of large language models (LLMs) throughout all areas of life, recent work suggests that interacting with LLM agents can be instrumental within the field of SLA and foreign language education, especially for reducing FLA. Related work also suggests that linguistic demands and task complexity can be predictors of FLA, implying that the use of demanding, complex language could lead to learners experiencing higher FLA. In this paper, we propose a novel multi-agent embodied conversational system that generates level-appropriate dialogue for English language learners. These levels are based on those defined by the Common European Framework of Reference for Languages (CEFR) to describe non-native listener and speaker proficiency. Using a “generate-evaluate-regenerate” loop with multiple LLM agents and a level classifier, it achieves a desired simplicity that is adaptive to the user’s proficiency level. We also share the results of a preliminary small-sample pilot study that tested this system with Japanese university students, to see whether it would yield lower FLA levels than an unsimplified embodied conversational agent. Analysis of conversational output showed that 87.4% of dialogue sentences generated by the proposed multi-agent system fell within one predicted CEFR level of the learner’s self-assessed proficiency, compared to 54.1% for the unsimplified agent. This suggests that the novel system is better able to produce output at an appropriate level for the learner. Though this study did not yield statistically significant evidence that the system reduces FLA levels in Japanese learners of English, likely due to a small sample size, it provides usability findings and culturally-informed design insights that will inform future study.
[HC-11] VisCanvas: A Node-based Interface for Exploratory Visualization Authoring with LLM s
链接: https://arxiv.org/abs/2607.21886
作者: Yuki Ueno,Bretho Danzy III,Zhuojun Jiang,Chris Bryan
类目: Human-Computer Interaction (cs.HC)
备注:
Abstract:Visual data analysis involves both open-ended exploration and targeted question answering. Visualization authoring tools support this process by enabling users to create visualizations for these tasks. With the rise of large language models, substantial effort has been devoted to developing visualization authoring tools that use natural language instructions. However, existing systems are typically based on a linear chat interface, which is not well suited to exploratory visual analysis workflows. In this paper, we introduce VisCanvas, a node-based interface for exploratory visualization authoring with LLMs. VisCanvas allows users to create, revise, branch, and merge visualizations in a non-linear way, enabling more efficient exploration of multiple analytical directions. We conducted a user study with 20 participants to evaluate the effectiveness of VisCanvas compared to a baseline chat-based interface. The results show that VisCanvas facilitates more diverse data interaction while maintaining performance levels (i.e., cognitive load and usability) that are indistinguishable from current prevailing methods. We then distill design principles for future AI-assisted visualization authoring environments. All supplemental materials required to reproduce the study are available at this https URL.
[HC-12] Filtering Offensive Content Changes Its Visibility but Not User Behavior: Two Randomized Controlled Trials with 200000 Users on Nextdoor
链接: https://arxiv.org/abs/2607.21853
作者: David J. Grüning,Matthew Katsaros
类目: Human-Computer Interaction (cs.HC); Social and Information Networks (cs.SI)
备注: 9 pages, 3 figures, 7 tables
Abstract:We investigate the effectiveness of interventions that reduce the visibility of offensive content on the local social platform Nextdoor. Content filtering – hiding or downranking offensive content that brushes against a platform’s rules without clearly breaking them – is deployed across virtually every major platform, yet almost no field evidence exists on whether it changes user behavior. We report two large-scale randomized controlled trials, each involving 100,000 users. Study 1 (2022) tested a report-triggered filter applied to comments in post threads and produced a modest 12% reduction in views of offensive comments; across eleven further measures of platform behavior we found no significant effects. Study 2 (2023-2024) remedied Study 1’s central limitation – a weak manipulation driven by slow, report-based eligibility – by proactively scoring posts and comments at creation with Google Jigsaw’s Perspective API and filtering them from the newsfeed. This produced a near-complete (95%) reduction in views of offensive posts, yet across thirteen further measures we again found no significant effects. Across two independent trials spanning different content types, filtering mechanisms, classifiers, and countries – and despite manipulation strength rising from 12% to 95% – filtering reliably reduced the visibility of offensive content without altering platform visitation, content consumption, or content production. These convergent null results provide rare field evidence on a ubiquitous intervention and underscore the complexity of effectively moderating online platforms.
[HC-13] Adaptive Driving Style for SAE Level-2 Driving Automation: Minimizing Preference Mismatch
链接: https://arxiv.org/abs/2607.21819
作者: Kumar Akash,Zhaobo Zheng,Teruhisa Misu,Vidya Krishnamoorthy,Mia Dong,Yuni Lee,Gaojian Huang
类目: Human-Computer Interaction (cs.HC); Systems and Control (eess.SY)
备注: Published in the 2026 American Control Conference (ACC), May 26-29, 2026, New Orleans, Louisiana, USA
Abstract:Driving style is a key factor in the comfort and acceptance of automated vehicle (AV) features. In SAE Level-2 automation, where the driver must supervise the system and remain ready to intervene, mismatches between the automation’s driving style and the driver’s preference can reduce trust and trigger takeovers. This paper proposes an adaptive driving-style control framework that minimizes such preference mismatch. In a driving-simulator study, we compare fixed, trust-based, and preference-based adaptation heuristics and analyze their effects on preference mismatch and trust. We then train a driving-preference prediction model and use it in an implicit adaptation policy that selects among bounded driving styles for upcoming events. A validation study shows that the predictive policy achieves equal or lower preference mismatch than comparison baselines, particularly when starting from a less defensive style, while also yielding higher average trust. The results provide a step toward developing human-aware driving automation that can implicitly adapt its driving style to the driver’s preferences.
[HC-14] AI-Integrated Scientific Inquiry: A Practice-Centered Vision for Science Education
链接: https://arxiv.org/abs/2607.21777
作者: Arne Bewersdorff,Matias Rojas,Xiaoming Zhai
类目: Human-Computer Interaction (cs.HC); Artificial Intelligence (cs.AI)
备注: Preprint
Abstract:Artificial intelligence (AI) has become part of scientific inquiry. Scientists use AI to observe and measure phenomena, to identify patterns in data, and to build models. As AI moves into scientific inquiry, it gains relevance for science education: students should learn how AI is changing scientific practices, ideally by engaging in AI-integrated scientific inquiry themselves. How to design such instruction, grounded in authentic scientific practice rather than taught as a standalone topic, remains an open question. In our vision, which we describe in this article, AI is treated as a set of scientific instruments that students use within the scientific practices described by the Next Generation Science Standards. Each instrument is a genuine scientific tool, pedagogically bounded: its controls are simplified while its core scientific function is preserved. The approach has two aims: engaging students in authentic scientific inquiry, and building an understanding of how AI is used in science and where it can mislead (discipline-based AI literacy, DAIL). In the article, we focus on the investigative core of inquiry, namely observing, analyzing, and modeling, and describe one exemplary AI instrument for each: computer vision for observing, clustering for analyzing, and generative modeling for modeling. We argue that every AI instrument in science education should carry a distinct reflection point that prompts critical evaluation of the AI instrument itself. Finally, we describe how agentic AI, operating across the whole inquiry rather than a single practice, could be represented, arguing that students should first build a foundational understanding of scientific inquiry and AI instruments before relying on agentic AI.
[HC-15] From Grasping to Speaking: Generative AI-Based Environment-Grounded VR Communication Training for Autistic Individuals
链接: https://arxiv.org/abs/2607.21769
作者: Ziming Li,Roshan L. Peiris
类目: Human-Computer Interaction (cs.HC)
备注: 14 pages, ASSETS2026
Abstract:Autistic individuals often face barriers in workplace communication, where soft skills are embedded within ongoing tasks and surrounding environment context, not in isolated verbal exchange. Recent work has introduced LLM-driven agents into VR-based communication training and proposed prompting schemas that let agents generate dialogue grounded in the VR environment and the user’s hand-based interactions. Building on this work, we explore how different levels of environmental grounding influence the training experience of autistic trainees and job coaches. We conducted an exploratory study with 9 autistic trainees and 7 job coaches across three modalities: conversation-only ©, conversation with environmental objects (C+O), and conversation with objects and grasp interactions (C+O+G). Usability and workload were comparable across modalities, while both trainees and coaches preferred the more interactive and environment-grounded condition (C+O+G). Participants viewed C+O+G as helpful for sustaining engagement and integrating communication practice into task performance. We discuss design considerations for more flexible, interactive, and workplace-relevant VR communication training for autistic individuals.
[HC-16] Bespoke Visual Assistance: What and How do Blind and Low-Vision People Create with Agent ic Programming?
链接: https://arxiv.org/abs/2607.21760
作者: Ellie Seehorn,Gene S-H Kim,Aziz Zeidieh,Ather Jammoa,Kun Lee,Aditi Shah,Jaylin Herskovitz,Anhong Guo,Venkatesh Potluri
类目: Human-Computer Interaction (cs.HC)
备注: 24 pages, 4 figures. To be presented at The 28th International ACM SIGACCESS Conference on Computers and Accessibility (ASSETS '26)
Abstract:AI-powered assistive technologies have long supported blind and low vision (BLV) people in everyday tasks, but they are general-purpose and often fall short of meeting complex, individualized, in-situ accessibility needs. Though agentic programming tools, like GitHub Copilot, have the potential to bridge this gap by lowering the technical barriers to building personal AT using natural language, the practical applicability of this creation paradigm has been unknown. We address this knowledge gap through a two-phase longitudinal co-design study with five tech-savvy BLV users using ProgramAT, an agentic programming tool that supports the creation, iteration, and testing of camera-based AT. Overall, co-designers created over 37 tools, with some addressing needs unmet by any existing commercial AT such as identifying Uber rides or interpreting hand gestures. Qualitative feedback from our co-designers and analysis of development logs surface BLV strategies for tool creation, along with key challenges including model capability limits, specification conflicts, and barriers to successful creation. We discuss recommendations to provide appropriate conversational scaffolding, community tool sharing capabilities, and support for specialized models and personal datasets for future agentic programming environments to empower BLV users to create bespoke visual assistance for themselves.
[HC-17] Co-design of LLM -based preference agents : participation may drive overtrust
链接: https://arxiv.org/abs/2607.21757
作者: Michael J. Fell
类目: Computers and Society (cs.CY); Artificial Intelligence (cs.AI); Human-Computer Interaction (cs.HC)
备注: 43 pages (18 main text plus supplementary material), 5 figures
Abstract:Large language models are increasingly used to simulate human preferences in research and practical applications, raising concerns about validation, misrepresentation, and exclusion. Co-designing agents with the people they represent is a promising way to address these concerns, but participation may also mask the problems it appears to solve. This paper explores that tension through a primarily qualitative study in which 12 participants co-designed personal preference agents in the domain of household energy, via a background survey, co-design interview, and validation survey. Participants engaged readily and mostly came to see their agents as representing them well. Independent validation, however, revealed mixed human-agent alignment, with agent responses markedly more homogeneous, decisive, and abstract than the human sample. I argue that participation and process transparency can act as an “overtrust engine” that promotes trust while concealing systematic misalignment with potential structural consequences at scale. I develop this as a core mechanism in participatory preference agent design, treating individual alignment not as a fixed state but as an enacted process.
[HC-18] Charting the Moral Universe: Capturing Virtues and Values of Data Visualization Practice
链接: https://arxiv.org/abs/2607.21732
作者: Chloe Hudson Prock,Enrico Bertini,Michael Correll
类目: Human-Computer Interaction (cs.HC)
备注:
Abstract:What do we value in our visualizations, and in the people who design them? Despite a growing body of work on critical data visualization, the conception of what it is to do ethical data visualization work can often be narrow (for instance, holding that our ethical duties are discharged merely by avoiding overtly lying or manipulating data), or entangled with potentially problematic implicit value structures (such as the assumption of the objectivity and neutrality of data, and so the designer’s role being merely the passive conveying of numbers as efficiently as possible). Yet, what it means to act ethically in data visualization is broad and multifaceted, and the virtues to which we should aspire as data visualization researchers and designers are worth explicating. We conducted an interview study with a broad spectrum of 20 experienced data visualization researchers, practitioners, and data artists to solicit their values and ethical considerations around doing visualization work. We report on a list of 68 values, organized into nine virtue clusters, that we encountered in our interviews. These virtues and values together describe a diverse space of matters of care and concern in data visualization: from the unease around the best use of visualization as a tool for persuasion, to the tightrope that visualization practitioners often walk between their professional responsibilities and their personal moral commitments. The virtues themselves, as well as our interviewees’ reflections on ethical practice, offer practitioners, researchers, and educators in data visualization a richer vocabulary for ethical reflection and provide a broader foundation for considering and applying visualization ethics.
[HC-19] A Systematic Survey on Image Description Techniques for STEM Domains
链接: https://arxiv.org/abs/2607.21611
作者: Marco Cardia,Letizia Angileri,Marina Buzzi,Giulio Galesi,Barbara Leporini
类目: Human-Computer Interaction (cs.HC); Artificial Intelligence (cs.AI)
备注: This is the Author’s Original Manuscript of an article accepted for publication in International Journal of Human-Computer Interaction, published by Taylor and Francis. The Version of Record is available at DOI https://doi.org/10.1080/10447318.2026.2668031
Abstract:The proliferation of visual data in Science, Technology, Engineering, and Mathematics (STEM) fields presents accessibility barrier for individuals with blindness or visual impairments. While recent advances in Artificial Intelligence (AI) offer new opportunities to generate textual descriptions of STEM images, the research landscape is fragmented and its impact on real users remains limited. This systematic survey examines 20 peer-reviewed studies on AI-based techniques for describing STEM visuals, with a specific focus on accessibility and human-computer interaction. Following the PRISMA methodology and a ROBIS-based risk-of-bias assessment, the review analyzes (i) the types of STEM visuals targeted, (ii) the AI and machine learning architectures employed, (iii) the datasets and evaluation metrics adopted, and (iv) the interaction modalities through which descriptions are delivered. The analysis reveals a shift from static, one-shot alt text toward interactive and multimodal systems that integrate conversational interfaces, keyboard navigation, and audio or haptic feedback. However, critical challenges persist, including factual inaccuracies and hallucinations, the scarcity of accessibility-first datasets co-designed with blind and low-vision users, and a heavy reliance on automatic text-overlap metrics that poorly capture perceived usefulness and trust. The survey concludes by outlining key research directions for HCI, emphasizing user-controlled verbosity, explainable and verifiable AI pipelines, and the integration of accessible description tools into mainstream STEM authoring and learning environments.
[HC-20] Natural Language Processing in Health Professions Education: A Scoping Review ALT
链接: https://arxiv.org/abs/2607.21605
作者: Javad Mohammad Alizadeh,Pegah Saebi,Mukesh Kumar Patel,Huanmei Wu
类目: Human-Computer Interaction (cs.HC)
备注: This manuscript has been accepted at the 14th IEEE International Conference on Healthcare Informatics (ICHI 2026)
Abstract:Natural language processing (NLP) and artificial intelligence (AI) are rapidly transforming health professions education, yet no scoping review has systematically mapped their applications across the full spectrum of health education contexts, including public health. Following the Arksey and O’Malley framework and PRISMA-ScR guidelines, this review synthesized evidence from 64 studies published between 2015 and 2026, identified through searches of PubMed, ERIC, IEEE Xplore, and Google Scholar. Seven thematic domains were identified: automated assessment, large language models (LLMs) as student-facing learning support, virtual patients and clinical simulation, curriculum analysis and program evaluation, personalized and adaptive learning, public health and health promotion education, and educator and institutional integration. Findings reveal significant technical promise, particularly in automated assessment, clinical simulation, and curriculum analysis, alongside persistent challenges including hallucination and accuracy concerns, algorithmic bias, data privacy, digital divide, overreliance, and absence of standardized outcome measures. Only four studies explicitly addressed public health education, representing a major evidence gap. This review provides a foundation for evidence-informed integration of NLP and AI across health professions education and highlights public health education as a priority area for future research and investment.
[HC-21] Analyzing Middle School Students Dialogue and Behaviors during Collaborative AI Chatbot Development Using Ordered Network Analysis
链接: https://arxiv.org/abs/2607.21603
作者: Shan Zhang,Andres Felipe Zambrano,Xiaoyi Tian,Yukyeong Song,Anthony F. Botelho,Kristy Elizabeth Boyer,Maya Israel,Shiyan Jiang
类目: Human-Computer Interaction (cs.HC); Artificial Intelligence (cs.AI)
备注:
Abstract:As Artificial Intelligence (AI) education has become a key component of K-12 curricula, activities such as designing and developing conversational agents are increasingly used as instructional practice. Prior work has primarily examined these activities by focusing on students’ learning outcomes or the quality of final AI artifacts, offering limited insight into the collaborative processes through which learning unfolds during AI system development. Although the AIED community has a long history of studying collaborative learning in STEM and Computing education, the emergence of AI learning environments in which students build AI systems presents new opportunities to understand how collaboration unfolds in AI education contexts. Grounded in these foundational works, the current study examines collaborative interaction among middle school students engaged in the design and development of an AI chatbot. Using Ordered Network Analysis of students’ dialogue and development actions, we characterize how collaboration is organized over time and how interaction patterns relate to chatbot quality and AI knowledge outcomes. Results reveal that higher-quality chatbots are associated with more integrated sequences linking explanation, testing, and refinement. Interaction patterns involving articulated reasoning and repeated testing and revision in response to chatbot output were also associated with stronger AI knowledge outcomes. These findings provide a process-oriented account of collaborative AI chatbot development and extend AIED research on collaborative learning processes to AI education contexts.
[HC-22] Control panels to clarify user intent with Large Language Models
链接: https://arxiv.org/abs/2607.21598
作者: Ben Shneiderman
类目: Human-Computer Interaction (cs.HC); Artificial Intelligence (cs.AI)
备注: 4 pages, 1 figure
Abstract:Typical user interfaces for Large Language Models present a blank prompt window that invites a natural language query by users, but offers little guidance. This paper proposes a visual control panel interface that would provide more cues to the semantics of prompt formation, enabling users to more easily express their intent. By emphasizing recognition over recall, control panels help users formulate more effective prompts that match their intent.
计算机视觉
[CV-0] Robot-Factored World Models via Robot Rendering
链接: https://arxiv.org/abs/2607.22535
作者: Byungjun Kim,Taeksoo Kim,Hyunsoo Cha,Hanbyul Joo
类目: Robotics (cs.RO); Computer Vision and Pattern Recognition (cs.CV)
备注: Project Page: this https URL
Abstract:Action-conditioned video world models predict future observations from an initial observation and an action signal. In robotics, actions influence future observations through two distinct processes: they are first realized into robot motion by the robot body and controller, and the scene then responds through contact and object motion. Conditioning directly on action commands asks the world model to learn the realization process itself, while conditioning on logged future states leaks the interaction outcomes it is meant to predict. We propose robot-factored world models, which move two robot-specific factors outside the world model. First, action realization: each command is rolled through the robot’s own controller and kinematics into a deployment-available nominal trajectory, a middle signal that avoids both action-realization learning and future-state leakage. Second, robot rendering: this nominal trajectory is rendered through the robot URDF, factoring the robot’s geometry, kinematics, and appearance out of the model and into explicit rendered robot geometry. To resolve depth ambiguity, we pair end-effector depth with scene depth, giving geometric cues for contact and occlusion beyond image-plane overlap. Together, camera-aware static RGB/depth context and rendered robot geometry form a shared visual world-model interface that stays consistent across viewpoints and robot embodiments, so the model sees the action only as visible robot geometry and learns how objects respond to it. Our experiments show that the rendered interface outperforms vector-conditioned baselines and generalizes to unseen robot embodiments at inference. We further demonstrate that our model generates robot manipulation videos from human demonstrations by retargeting and rendering the hand motion as robot geometry.
[CV-1] SM4RT: Learning Structured Motion Geometry for 4D Reconstruction
链接: https://arxiv.org/abs/2607.22534
作者: Shing Ho J. Lin,Wenzhao Zheng,Dong Zhuo,Yuqi Wu,Jie Zhou,Jiwen Lu
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI); Robotics (cs.RO)
备注: Code is available at: this https URL
Abstract:Geometry Foundation Models (GFMs) have substantially advanced monocular 3D reconstruction, yet extending this capability to 4D dynamic understanding remains a fundamental challenge. Most existing motion perception methods (e.g., sparse tracking, dense point-wise flow) treat motion as independent point-wise displacements, ignoring the structured nature of physical motion. However, real-world objects usually obey rigid-body kinematics, and points thus usually move collectively, not in isolation. Motion itself possesses geometric structure: physical objects undergo a set of rigid-body transformations governed by SE(3), rather than unstructured point-wise displacements. Building on this insight, we propose SM4RT, a Structured Motion 4D Reconstruction Transformer for end-to-end 3D reconstruction and structured motion perception. SM4RT introduces Structure-of-Motion to represent scene dynamics, where scene motion is decomposed into a compact set of motion bases, each represented as a temporal sequence of 6D twists in SE(3). Dense scene motion is then recovered by sparse, time-shared per-pixel assignment weights over these bases, ensuring points on the same object share a common rigid-body motion trajectory. SM4RT introduces a parallel motion geometry encoder and decoder that jointly infer 3D geometry, world-coordinate motion, and scene kinematic structure in a single forward pass from monocular RGB video. SM4RT achieves strong motion reconstruction performance while preserving the geometric structure of scene motion.
[CV-2] wins: Learn to Predict Unified Representations with Focal Loss ICML2026
链接: https://arxiv.org/abs/2607.22531
作者: Kaixiong Gong,Xin Cai,Bin Lin,Hao Wang,Yunlong Lin,Mingzhe Zheng,Bohao Li,Jian-Wei Zhang,Miles Yang,Zhao Zhong,Liefeng Bo,Xiangyu Yue
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: ICML 2026. Code: this https URL
Abstract:Unified multimodal models seek a shared visual token space that supports both multimodal understanding and image generation. Discrete methods unify the interface via a shared codebook, whereas continuous pipelines often rely on two disparate representations – semantic features (e.g., ViT) for understanding and low-level latents (e.g., VAE) for synthesis – resulting in mismatched latent spaces. We propose Twins, a unified continuous token space formed by channel-wise concatenating ViT and VAE features on the same token grid, so the sequence length is unchanged and attention cost does not increase. However, jointly modeling Twins in a Diffusion Transformer exposes a severe optimization imbalance: the model fits the ViT component well but struggles to match the VAE latent distribution. We trace this imbalance to three sources of heterogeneity: frequency bias, intrinsic dimensionality, and condition-aligned vs condition-independent uncertainty. To address it, we adapt a focal regression objective for flow matching that upweights large-error VAE dimensions, better balancing optimization across the ViT and VAE components. On ImageNet, this yields up to 10.57 gFID gain over naive MSE loss without classifier-free guidance. Twins also performs competitively on multimodal understanding benchmarks and improves reconstruction fidelity, narrowing the gap between understanding- and generation-oriented representations.
[CV-3] CARA: Concept-Aware Risk Attention for Interpretable Collision Anticipation
链接: https://arxiv.org/abs/2607.22494
作者: Zhishan Tao,Ruoyu Wang,Yucheng Wu,Enjun Du,Yilei Yuan,Sherwin Ho,Yue Su,Jinbo Su,Yi Hong
类目: Multimedia (cs.MM); Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted to ACM Multimedia 2026(Oral)
Abstract:Collision anticipation in autonomous driving requires not only accurate early warnings but also interpretable reasoning about what risk factors are being tracked and how risk evolves over time. Existing methods fall short in this regard: feature-driven models are opaque, post-hoc explanations often lack fidelity, and concept-based methods are mostly designed for static recognition rather than dynamic driving scenes. We propose CARA (Concept-Aware Risk Attention), an intrinsically interpretable spatio-temporal framework for collision anticipation. CARA derives domain-grounded risk concepts from accident narratives, aligns them with video frames via vision-language similarity, and organizes them into evolving concept trajectories. These trajectories provide explicit risk evidence that guides spatial attention, temporal attention, and anticipation, allowing semantic concepts to directly influence both where the model attends and how it predicts risk over time. By treating semantic risk factors as dynamic intermediate evidence rather than auxiliary post-hoc explanations, CARA tightly couples interpretability with the predictive process. Extensive experiments on three benchmarks show that CARA consistently improves anticipation accuracy and warning earliness over strong baselines, while providing sparse and semantically grounded concept evidence.
[CV-4] Optimal Transport Image Representation and Deep Covariance Alignment (CORAL) for Control Valve Stiction Detection
链接: https://arxiv.org/abs/2607.22486
作者: Seshu K. Damarla
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 7 pages, 6 figures, 6 tables
Abstract:Control valve stiction is a common cause of unwanted oscillations and poor control-loop performance in industrial processes. Data-driven methods can automatically detect stiction, but models trained purely on simulated data often struggle to generalize to real industrial control loops due to domain shift. To bridge this gap, this work propose a novel stiction detection methodology that combines optimal transport (OT) imaging technique with deep correlation alignment (Deep CORAL) algorithm. Closed loop signals: controller output and process variable are converted into two-dimensional OT images. These images capture the dynamic behaviour of control loops. The proposed methodology includes a convolutional neural network (CNN) encoder (or feature extractor) trained to learn domain-invariant representations by optimizing a combined objective: a cross-entropy loss on labeled simulation data and a Deep CORAL (covariance-alignment) loss between simulation data and unlabeled target-domain industrial data. Downstream classifiers trained on the domain-invariant target features were evaluated on an independent test set of 20 benchmark loops from industrial stiction data benchmark. The proposed methodology successfully diagnosed 18 out of the 20 loops and achieved100% recall across all 13 stiction cases, an accuracy of 90.00% and an F1-score of 92.86%. Compared to standard baseline approach (hand-crafted features-based method), the proposed methodology significantly mitigates domain shift, providing robust, highly reliable stiction detection for real-world industrial control loops.
[CV-5] Deformable Triangle Splatting: Flexible Primitives for Real-Time Radiance Field Rendering ECCV2026
链接: https://arxiv.org/abs/2607.22446
作者: Oriol Jiménez-Ayguadé,Antonio Agudo
类目: Computer Vision and Pattern Recognition (cs.CV); Graphics (cs.GR)
备注: Accepted at ECCV 2026. Project page: this https URL
Abstract:Recent radiance field methods represent scenes with 2D primitives that offer surface alignment and efficient rasterization, from Gaussian disks to triangles, yet all rely on convex boundaries: curved and concave structures demand excessive primitives. We introduce Deformable Triangle Splatting, which augments each triangle with K control points per edge, each parameterized by a single learnable scalar displacement that shifts the boundary inward or outward, enabling non-convex shape representation while preserving the three base vertices that define the 3D plane. To render these non-convex primitives differentiably, we design a rasterization pipeline in the triangle’s barycentric coordinate space, ensuring view-consistent rendering. A winding number test determines whether each pixel lies inside the deformed primitive, and a window function controlled by two learnable parameters, sharpness and corner smoothness, together with a per-primitive scalar opacity, produces the smooth opacity transition from interior to boundary. Validation is done in a variety of real-world scenes, outperforming recent works based on non-volumetric primitives in terms of visual quality and versatility while still achieving competitive rendering efficiency.
[CV-6] SceneActBench: Can Agents Act on the 3D Scenes They See?
链接: https://arxiv.org/abs/2607.22393
作者: Yifei Zhao,Xiangxin Zhou,Wenhao Yang,Jiaqi Tang,Pu Jian,Huanjin Yao,Jiarui Yao,Haowei Lin,Chunchao Guo,Zhuo Chen,Wenkai Lyu,Jianzhu Ma,Xueqian Wang,Wenxi Zhu
类目: Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Vision-language model (VLM) agents increasingly use tools to act on 3D scenes rather than only describe them. Existing 3D benchmarks score textual responses or single-object operations, leaving agent action on complete multi-object 3D scenes under evaluated. We present SceneActBench, a benchmark for visually conditioned action across five 3D tasks under a unified agent-environment loop. Given PNG images or sampled video frames and, where applicable, supplied 3D assets, an agent acts on a 3D environment. We evaluate each final output against hidden ground truth with task-specific geometric metrics. SceneActBench comprises five tasks built from 210 source instances, yielding 520 task cases including paired input conditions. Every task runs through one fixed agent loop to keep the comparison fair. Across eleven proprietary VLM configurations, Overall scores span 38.6-50.2, and none performs consistently well across tasks. We further analyse where and how failures manifest.
[CV-7] Correlation-Aware and Gaussianity-Preserving Robust Latent Angular Watermarking for Diffusion Models
链接: https://arxiv.org/abs/2607.22386
作者: Yebin Zheng,Haonan An,Guang Hua,Zhiping Lin,Yuguang Fang
类目: Computer Vision and Pattern Recognition (cs.CV); Cryptography and Security (cs.CR)
备注:
Abstract:Latent domain watermarking for diffusion models embeds watermarks directly into the latent prior, enjoying non-intrusiveness to model parameters and seamless integration with the generation process. However, due to the violation of latent Gaussianity or sensitivity to normal and malicious perturbations during latent inversion, existing methods are prone to watermark detection or removal attacks. A further overlooked problem is the violation of the i.i.d. latent condition after watermarking, which leads to latent correlation degradation and generation fidelity loss. Although this has been externally measured by FID, the internal correlation structure has yet to be rigorously characterized. To address the above issues, and motivated by the rotation-invariant property of isotropic Gaussian, we propose \textitLatent Angular Watermarking (LAW), which encodes watermark bits as antipodal angles ( \pm\pi/2 relative to a reference pair) between disjoint pairs of latent elements while preserving the Gaussianity. The antipodal ( \pi -separation) encoding maximizes geometric separation between bit values, and we prove that the decoding angular-error variance is proportional to the norm of the latent pair, i.e., \operatornamevar(\Delta\phi) \propto 1/\rho^2 . We further propose a magnitude-driven variant, LAW-M, which anchors watermark bits in the most geometrically stable latent dimensions, yielding additional robustness gains. Theoretically, we provide a rigorous characterization of the induced correlation degradation, deriving in closed form the autocorrelation structure of the watermarked latent and proving that correlations are confined to a sparse, structured set of off-diagonal elements with fixed \pm\pi/4 values.
[CV-8] IR275K: A Benchmark for Infrared Multi-Frame Super-Resolution Toward Efficient Remote Sensing
链接: https://arxiv.org/abs/2607.22380
作者: Jie Deng,Heyang Wang,Changxin Wang,Junkai Shen,Hongyi Chen,Zhiping He,Hongxing Qi,Xudong Zhang,Jianyu Wang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Efficient processing is becoming increasingly important in infrared remote sensing, where satellite constellations produce large volumes of observations under constrained detector resolution, power, and downlink bandwidth. Multi-frame super-resolution (MFSR) offers a software-based route to spatial enhancement, but its evaluation in infrared sensing remains fragmented across private datasets and ad-hoc protocols. Existing benchmarks do not explicitly capture the thermal contrast, sensor noise, weak texture, and platform-induced frame-to-frame variation that characterize infrared video. We introduce IR275K, a curated benchmark containing 594 infrared video sequences and 275,196 frames. It provides sequence-level train/validation/test splits and a reproducible X4 evaluation protocol. As an initial architectural probe, we further evaluate CGMamba, a lightweight state-space model with 10.90M parameters and 112.14G FLOPs. CGMamba combines 2D rotary position encoding (2D~RoPE) with center-guided cross-Mamba (CGCM) fusion for implicit multi-frame reconstruction. It achieves 33.19dB PSNR, outperforming infrared single-image super-resolution references by 0.35–0.52~dB at substantially lower computational cost. Ablation results show that removing 2D~RoPE from CGCM causes a 1.53dB drop and severe grid-like artifacts. This indicates that explicit spatial anchoring is critical for stabilizing SSM-based cross-frame gating under infrared conditions. IR275K provides a reproducible foundation for accuracy–efficiency evaluation of infrared MFSR methods, while the architectural analysis offers a concrete starting point for spatially aware SSM design under resource-constrained infrared sensing. Dataset and evaluation resources are available at: this https URL.
[CV-9] Active few-shot segmentation by reinforcing data selection MICCAI MICCAI2026
链接: https://arxiv.org/abs/2607.22371
作者: Chenlan Zhao,Benny Wong,Timothy F. Lundberg,Ahmed M. Elsayed,Abdallah Aljarkas,Hamad A. Aljamaan,Lynn Karam,Qianye Yang,Yipeng Hu,Claire C. Villette,Shaheer U. Saeed
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted at EMA4MICCAI 2026 - The 2nd MICCAI Workshop on Efficient Medical AI
Abstract:Few-shot learning enables medical image segmentation models to adapt to new tasks using only a small number of labelled examples. However, adaptation performance depends strongly on which examples are selected for the support set. Effective support sets should capture relevant variation within the target domain and be informative for adaptation, with constituent samples providing complementary information. Despite this, existing active data selection approaches largely prioritise samples individually and do not explicitly account for interactions between examples. In this work, we propose a reinforcement learning framework for support-set selection in few-shot medical image segmentation, enabling support sets to be optimised jointly rather than through independent sample scoring. Given a pool of unlabelled candidate images, an agent directly predicts a support set that maximises downstream segmentation performance. Experiments on a cross-institutional pelvic MRI dataset demonstrate improvements over random selection and current state-of-the-art methods. Our findings highlight the importance of support-set complementarity for effective adaptation and demonstrate the potential of reinforcement learning for optimising adaptation sets.
[CV-10] SiPhy: Single-Image Physical Property Reasoning ECCV2026
链接: https://arxiv.org/abs/2607.22355
作者: Hoang Le,Joonwoo Kwon,Elkhan Ismayilzada,Yufei Zhang,Zijun Cui
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: Accepted to ECCV 2026 (main track)
Abstract:Inferring physical properties such as mass, stiffness, and elasticity from a single image is essential for simulation and embodied AI, yet most existing approaches rely on multi-view reconstruction or physics-based supervision. We introduce SiPhy, a unified framework for single-image physical property reasoning that aligns 3D-aware visual cues, depth with language-based material knowledge. From one RGB image, SiPhy samples pseudo-voxel points, extracts CLIP features, and grounds them to material candidates proposed by a VLM. A part-based contrastive aggregator enforces region consistency, while a heaviness-aware refinement improves thickness and volume estimation for dense objects. Across ABO-500, MVImgNet-100, and PhysXNet-100, SiPhy achieves state-of-the-art single-image performance, surpassing multi-view reconstruction methods by improving mass MnRE by up to 93% (vs. PUGS), reducing density MAE by 35.5% (vs. NeRF2Physics), and lowering Young’s modulus error by 23.5%. We further validate SiPhy on real hand-object interaction datasets, demonstrating its potential as a data annotation engine for physical understanding from single-view imagery.
[CV-11] me-Reversed Imaging: A Multimodal Benchmark and Framework for Reconstructing Past Human-Environment Interactions
链接: https://arxiv.org/abs/2607.22352
作者: Jorge Bacca,Kebin Contreras,Luis Toscano-Palomino,Mauro Dalla Mura
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI); Image and Video Processing (eess.IV)
备注:
Abstract:We introduce time-reversed imaging, a new paradigm that infers what just happened in a scene from fading multimodal traces. Instead of extrapolating or interpolating video frames, our goal is to infer past human-environment interactions from residual physical imprints observable in thermal, ultraviolet, and visible spectra. To study this problem, we present TRACE-HEI, the first proof-of-concept dataset for time-reversed imaging, containing synchronized tri-modal video sequences of actions such as sitting, touching, moving objects, and liquid spills, captured across diverse materials and recorded up to three minutes after contact. To establish the benchmark, we propose a multimodal inference approach that extracts structured textual descriptions of detected traces and uses them to constrain a vision-language-guided diffusion model for reconstructing plausible past frames. Experiments show that inferring recent events from fading traces is challenging but feasible when complementary modalities reduce solution ambiguity. This work defines the first computational and experimental foundation for time-reversed imaging, bridging vision, physics, and generative reasoning, and opening new directions for scene understanding beyond instantaneous observation.
[CV-12] SLIP: Segmentation with Low-latency Interactive Prompting for 3D Medical Images
链接: https://arxiv.org/abs/2607.22332
作者: Baptiste Podvin,Alexandre Ancel,Flavio Milana,Chiara Innocenzi,Davide Arrigo,Federico Espinola Schulze,Guido Torzilli,Jacques Marescaux,Daniel George,Alexandre Hostettler,Toby Collins
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Interactive deep image segmentation enables efficient medical image annotation by iteratively refining predictions from user prompts, such as positive and negative clicks. Recent patch-based methods, including nnInteractive, achieve strong segmentation performance but remain limited in annotation workflows by high interaction latency, limited responsiveness to successive interactions, and the lack of support for reversible prompting. Furthermore, evaluation relies predominantly on simulated rather than controlled real-user interaction studies. We present SLIP, an end-to-end trainable framework for interactive 3D medical image segmentation that decouples image encoding from prompt-guided refinement. Image features are computed once and reused, while a lightweight patch memory bank maintains an interaction-aware segmentation state shared across patches. This representation enables prediction updates by propagating interaction context throughout the image, supports reversible prompting without recomputing image features, and substantially reduces interaction latency. By separating image representation from interactive reasoning, SLIP remains compatible with a wide range of image encoders. We train a single SLIP model for general interactive segmentation across diverse anatomical structures and imaging modalities. Beyond standard simulated evaluation, we conduct a controlled prospective user study comparing manual segmentation, nnInteractive, and SLIP across three clinical annotation tasks, six expert participants, and subjective usability measures, addressing the limited human validation of interactive segmentation methods. SLIP achieves SOTA interactive segmentation performance across 13 public datasets while providing lower interaction latency, greater responsiveness, support for reversible prompting, and higher user preference than existing approaches.
[CV-13] Geometric 2D Scene Graph Generation
链接: https://arxiv.org/abs/2607.22325
作者: Christoph Jahn,Urs Waldmann,Bastian Goldluecke
类目: Computer Vision and Pattern Recognition (cs.CV); Robotics (cs.RO)
备注: Accepted at the 18th International Conference on Agents and Artificial Intelligence ICAART 2026, Marbella, Spain
Abstract:In production processes for consumer products, assembly instructions are essential not only for planning but also for executing the production process. Likewise in robotics, it is crucial for an assembly robot to understand how components fit together and can be assembled. To facilitate these tasks, we contribute a method for constructing scene graphs to represent and characterize assembly relationships between components. Our approach does not rely on semantic data and is capable of handling a very small dataset. To realize this, the output of a Faster R-CNN model is used to create geometric representations, which are then processed by a transformer architecture to generate an adjacency matrix. This matrix serves as input to a Siamese network that uses message passing based on an attentional graph convolutional network (aGCN) architecture to characterize the connections between the components. We validate our method on a study dataset of toy model components which can be assembled into transportation vehicles.
[CV-14] fMRI2Face: A Full-HD fMRI-Video Dataset and Geometry-Guided Neural Decoding Framework for Dynamic Human Face Reconstruction
链接: https://arxiv.org/abs/2607.22302
作者: Jingyang Huo,Xiangru Huang,Chentao Shen,Yikai Wang,Yun Wang,Jianxiong Gao,Shihao Jin,Yanwei Fu,Jianfeng Feng
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Reconstructing dynamic human faces from brain activity provides a powerful way to study how the mind perceives identity, expression, and facial motion. However, progress in fMRI-based face decoding has been limited by scarce controlled, high-resolution neural datasets and by methods that struggle to recover both identity-specific appearance and time-varying facial dynamics. We present fMRI-Face, the first fMRI dataset paired with controllable full-HD digital human facial videos rendered at 1920 \times 1080 resolution. During scanning, participants watched photorealistic, background-free facial videos with controlled identity, expression, and head pose, while fMRI activity was recorded. The resulting dataset contains 62,856 paired fMRI-video samples, providing a structured resource for studying dynamic face perception and reconstruction. Building on this dataset, we propose fMRI2Face, a geometry-guided neural video decoding framework for reconstructing facial videos from fMRI signals. fMRI2Face derives two complementary neural controls from brain activity: Brain-derived Appearance Context, which captures global identity-related visual attributes, and Morphable 3D Facial Control, which provides explicit geometry-aware guidance for pose, expression, and non-rigid facial dynamics. These controls are integrated through Neural-Controlled Video Diffusion with auxiliary latent completion, enabling high-fidelity facial video reconstruction directly from brain activity. Experiments show that fMRI2Face consistently improves reconstruction fidelity, identity preservation, facial geometry, and motion consistency over representative neural decoding baselines. Together, fMRI-Face and fMRI2Face establish a controlled platform for studying dynamic face perception and provide a new benchmark for fMRI-based digital human reconstruction.
[CV-15] RadSight: Towards Perceptually Reliable Multimodal Radiology Image Understanding
链接: https://arxiv.org/abs/2607.22293
作者: Jianqin Liu,Weiwei Cao,Wanxing Chang,Ruifeng Yuan,Bowen Shi,Zhilin Zheng,Xianjie Zhang,Ling Zhang,Peng Wang,Jianpeng Zhang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Medical multimodal large language models (MLLMs) are increasingly expected to perform complex image understanding tasks, yet their reliability is often compromised by frequent errors in visual interpretation. To systematically trace these failures, we traverse the hierarchy from high-level clinical tasks down to fundamental visual perception. We therefore introduce Perception-Bench, a large-scale benchmark comprising 1.13 million samples that assesses medical MLLMs across six dimensions: attribute judgment, spatial grounding, spatial understanding, disease prediction, anomaly detection, and report generation, spanning both 2D and 3D radiology images. Our analysis on Perception-Bench reveals that existing MLLMs lack the ability to capture even the most basic lesion attributes, such as location, size, and density. This inability to ground clinical outputs in primary visual evidence reveals that the models’ diagnostic unreliability is rooted in a critical but overlooked bottleneck in low-level visual perception. Motivated by this, we propose RadSight, a perception-driven MLLM built upon a dual 2D/3D encoder architecture that preserves native imaging spatial structures. RadSight formulates medical image understanding as a four-stage progressive process: visual-language alignment, fine-grained visual perception, clinical diagnosis, and diagnostic interpretation. The model is trained on an 8.37 million perception-oriented corpus using progressive curriculum learning. On Perception-Bench, RadSight consistently outperforms existing MLLMs across all six evaluation dimensions, with particularly strong gains in spatial grounding and clinical diagnosis. It also achieves consistent improvements on public 2D and 3D medical benchmarks, further demonstrating that robust low-level visual perception is a critical foundation for reliable clinical understanding. Code and model will be publicly available.
[CV-16] Class-Balanced Softmax: A Bayes Theory-Based Method for Long-Tailed Recognition
链接: https://arxiv.org/abs/2607.22258
作者: Yi-Hang Zhu,Rajeev Raman,Shiqi Su,Jianyuan Sun,Xinyu Yang,Nan Xing,Huiyu Zhou
类目: Machine Learning (cs.LG); Computer Vision and Pattern Recognition (cs.CV)
备注: 39 pages, 8 figures, under review in a journal
Abstract:Deep learning models using traditional softmax classifiers have achieved remarkable success in various classification tasks. However, their performance degrades significantly on imbalanced datasets. Although Balanced Softmax is widely adopted as a state-of-the-art rebalancing method, it possesses inherent limitations, such as yielding disproportionately lower testing accuracy for tail classes. To mitigate these shortcomings, we propose the Class-Balanced Softmax (CBS). Rooted in a theoretical Bayesian framework and a heuristic power-law assumption, the CBS is a simple logit adjustment that is computationally inexpensive and easily integrated into existing pipelines. Furthermore, we characterise a fundamental phenomenon in models trained on imbalanced data, termed the preference issue, wherein models exhibit higher training error and a larger generalisation gap for classes with limited data. To quantify this issue, we introduce a novel metric and demonstrate that CBS effectively mitigates the preference issue. Extensive experiments on large-scale benchmarks show that CBS is highly scalable and outperforms existing methods, including Balanced Softmax.
[CV-17] From level set evolution to threshold optimization: A grayscale level set framework for image segmentation
链接: https://arxiv.org/abs/2607.22255
作者: Xingkai Li,Jiebao Sun,Fanghui Song,Zhichang Guo
类目: Computer Vision and Pattern Recognition (cs.CV); Mathematical Physics (math-ph)
备注:
Abstract:The segmentation of multiple degradations has been a challenging problem in the field of image segmentation. Existing level set approaches commonly adopt a length regularization term to constrain the geometric shape of the segmentation contour. However, the introduction of the length term often results in numerical instability and high computational cost. In this paper, we show that the length term is not essential under certain smoothness constraints, and theoretically prove that the presence of the length term affects the property of |\nabla \phi|=1 . Based on the finding, we define a class of smooth images, construct the grayscale level set, and propose a fast segmentation framework for degraded images, such as heavily noisy images and intensity inhomogeneous images. The framework transforms PDE evolution into one-dimensional threshold search, which has significant advantages in computational speed, especially on large-scale images. Experiments validate the segmentation performance of the proposed framework on various degraded images.
[CV-18] Agent HOI: Multi-Agent Reasoning for Human-Object-Interaction Video Generation via Implicit Representation Alignment
链接: https://arxiv.org/abs/2607.22241
作者: Ziyao Huang,Shunkai Li,Juan Cao,Chenyu Li,Youliang Zhang,Zixiang Zhou,Cong Wang,Yuan Zhou,Qinglin Lu,Fan Tang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Recent advances in video diffusion models have spurred interest in human-object interaction (HOI) video generation, which demands fine-grained control over interaction logic beyond single-subject animation. However, existing HOI methods rely heavily on explicit motion control, limiting scalability and generalization across diverse objects and interactions. In this study, we propose AgentHOI, a text-driven HOI video generation following a thinking-before-generation framework that bridges the gap between high-level textual intent and physical execution through multi-agent reasoning over perception, interaction, and motion planning. Building upon the generated interaction plans, we further strengthen text-driven motion understanding. We introduce an implicit text-motion alignment strategy that distills text-to-motion priors into the video diffusion model, enabling robust HOI synthesis without explicit motion inputs at inference. Experiments show that AgentHOI significantly improves interaction naturalness, object appearance preservation, and adherence to complex textual instructions across challenging object-centric scenarios such as wearing and riding. The code is available at this https URL.
[CV-19] RaM-VSR: Importance-Aware Token Routing and Merging for One-Step Diffusion Video Super-Resolution
链接: https://arxiv.org/abs/2607.22231
作者: Sicheng Gao,Zhuyun Zhou,Yixuan Liu,Tong Shen,Zongwei Wu,Radu Timofte
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: 18 pages, 7 figures
Abstract:Video super-resolution (VSR) using large-scale Diffusion Transformer (DiT) priors achieves exceptional perceptual quality but is often impractical due to the quadratic computational cost of processing dense spatio-temporal token sequences. Existing efficiency-oriented methods risk irreversible detail loss and temporal flickering, a vulnerability especially pronounced in one-step diffusion models. To address this, we propose TRaM-VSR, a Token Routing and Merging framework for adaptive token allocation, leveraging both context-aware video priors and network-level priors. First, token importance is estimated by fusing motion-sensitive temporal cues with semantic text similarity, isolating dynamic objects and structural boundaries. Next, this importance is further calibrated and adjusted by an offline planner to guide routing across optimally grouped network blocks. Technically, within each routed group, structurally critical tokens are processed in a high-fidelity local stream, while less informative tokens are aggregated into a compact global stream, both modulated by network depth and aligned with the multigranular nature of diffusion models. Extensive experiments show that TRaM-VSR accelerates inference significantly while preserving state-of-the-art reconstruction quality and robust temporal consistency. The code is available at this https URL.
[CV-20] Deep Convolutional Large-Margin ell_p-SVDD for Visual Anomaly Detection
链接: https://arxiv.org/abs/2607.22212
作者: Alireza Dastmalchi Saei,Shervin Rahimzadeh Arashloo
类目: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注:
Abstract:Visual anomaly detection requires adaptive representations and reliable decision boundaries, particularly when anomalous training samples are scarce and class distributions are highly imbalanced. Classical kernel-based methods yield principled geometric decision regions but typically operate on fixed features, while deep detectors learn task-specific representations but often fail to provide an explicit margin-aware kernel boundary. In this study, we propose DLM-SVDD, a deep large-margin novelty-detection framework that jointly learns convolutional features and an explicit kernel-based decision boundary. By drawing on the large-margin \ell_p -Support Vector Data Description ( \ell_p -SVDD) approach, the proposed method performs explicit margin maximization and nonlinear slack penalization while adapting the representation to the target task. To train the proposed model, we present an optimization scheme that alternates between a Frank–Wolfe–based update of the convex dual boundary and a CNN update step operating on a smooth margin-violation loss induced by the recovered boundary. To improve scalability, we analyze the efficiency–accuracy trade-offs for different kernel approximation strategies, deriving practical propositions for large-scale anomaly detection. Extensive experiments on multiple standard benchmarks show consistent performance improvements over the baseline and strong overall performance compared with state-of-the-art methods while illustrating that the proposed joint representation–boundary learning scheme remains effective under severe imbalanced class distributions. Subjects: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG) Cite as: arXiv:2607.22212 [cs.CV] (or arXiv:2607.22212v1 [cs.CV] for this version) https://doi.org/10.48550/arXiv.2607.22212 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[CV-21] Filling Before Advancing: Capability-Gap-Driven Post-Training for Scenario-Specialized Remote Sensing MLLM s
链接: https://arxiv.org/abs/2607.22205
作者: Yuheng Zong,Minghua Wang,Xin Zhao,Zhi-Hui Zhan,Antonio Plaza,Jon Atli Benediktsson
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:
Abstract:Remote sensing multimodal large language models (RS-MLLMs) have improved general aerial-image understanding. However, Earth observation applications require fine-grained scenario specialization, constrained by scarce high-quality scenario data and incomplete capability coverage. We formulate this adaptation as a capability-gap-driven post-training problem and propose filling before advancing (FBA). Rather than relying on single-stage supervised fine-tuning (SFT) over target-domain samples, FBA first fills prerequisite capability gaps before advancing toward scenario specialization. We instantiate FBA for coastal harbor understanding, a representative multi-source scenario, by constructing CPRS (Coastal-Port Remote Sensing), a three-layer supervision dataset coupled with three ordered stages: (1) RS semantic anchoring for overhead-view visual-language alignment; (2) domain-bridge convergence for shared RS priors across target and bridging scenarios under different modalities; and (3) evidence-grounded scenario tuning for downstream performance. We construct HarborEval, an eight-track diagnostic benchmark covering perception, spatial understanding, robustness, and generation. Under comparable training budgets, HarborEval increases from 57.95 with Direct-SFT to 70.29 with FBA on LLaVA-v1.5, and from 81.09 to 83.37 on Qwen3-VL. FBA also outperforms Collapsed-SFT and leads on harbor-related VRSBench/RSVQA subsets and OpenEval. Stage-wise and role-replacement analyses validate progressive gap filling and stage-specific roles. Public examples and release updates for CPRS, HarborEval, code, and trained weights are available at this https URL.
[CV-22] LayoutLite: Token-Level Implicit Layout Analysis for Efficient Document OCR
链接: https://arxiv.org/abs/2607.22200
作者: Xudong Liu,Bicheng Wan,Yulin Jin
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 11 pages, 7 figures
Abstract:End-to-end OCR systems based on vision-language models have achieved strong performance in complex document OCR, but their efficiency is limited by the large number of visual tokens produced from document images. Many of these tokens correspond to blank margins or visually redundant regions, yet directly applying generic visual token compression methods may remove OCR-critical fine-grained details. In this paper, we propose LayoutLite, a lightweight plug-and-play module for efficient document OCR. Instead of relying on explicit document layout detection, LayoutLite performs implicit layout analysis at the token level between the vision encoder and the language decoder. It aggregates multi-layer visual representations from the vision encoder, and predicts an importance score for each visual token with a lightweight scoring network. Low-information tokens are then removed before entering the language decoder while preserving the original spatial positional information of retained tokens. To train LayoutLite without human annotations, we cast token selection as a reinforcement learning problem and optimize it with a group-relative policy optimization objective driven by OCR output consistency, together with an auxiliary layout supervision signal to stabilize training. Experiments on OmniDocBench demonstrate that LayoutLite can substantially reduce visual token length and inference cost with negligible degradation in recognition quality. We further evaluate LayoutLite on two OCR-specialized VLMs, FireRed-OCR and Logics-Parsing-V2. Under up to 50% token compression, LayoutLite preserves almost the same score on both models while reducing prefill latency, FLOPs, and KV cache memory by over 40%, with only a small additional inference overhead. These results show that token-level implicit layout analysis is an effective and practical approach for accelerating VLM-based OCR systems.
[CV-23] Bowel Obstruction Detection and Localization on Abdominal CT with Deep Learning
链接: https://arxiv.org/abs/2607.22173
作者: Moritz Vandenhirtz,Andrea Agostini,Dana Belde,Mélanie Roschewitz,Ismaiel Chikh Bakri,Tilo Niemann,André Euler,Julia E Vogt
类目: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注:
Abstract:Bowel obstruction is a common and potentially life-threatening gastrointestinal condition. In the face of rising diagnostic workloads, the automated diagnosis of bowel obstruction on CT scans supports radiologists by accelerating detection and improving patient outcomes. In this work, we propose a deep learning framework with a multi-task objective that jointly detects bowel obstruction and localizes its transition zone. Additionally, we extend the method with an inherently interpretable classification method that locates the suspected transition point within a slice. It does so by learning a probabilistic selection mask that faithfully bases the classifier’s prediction solely on a small image region. The proposed method is evaluated on an internal dataset comprising 1,427 abdominal CTs. Here, the model achieves an obstruction detection test accuracy of 93% and a Hit@10 transition zone localization of 95%. As the first method to reliably localize the transition zone, this marks a significant step towards the automated identification of this critical clinical landmark.
[CV-24] JustDepth: Real-Time Radar-Camera Depth Estimation with Single-Scan LiDAR Supervision
链接: https://arxiv.org/abs/2607.22172
作者: Wooyung Yun,Dongwook Kim,Soomok Lee
类目: Computer Vision and Pattern Recognition (cs.CV); Robotics (cs.RO)
备注: Project page: this https URL
Abstract:Accurate yet low-latency depth is essential for radar-camera perception in autonomous systems. Cameras provide rich appearance but lack metric scale, whereas automotive radar offers metric range but is sparse and noisy. Many pipelines are multi-stage or depend on auxiliary annotations, increasing latency and limiting portability. We introduce JustDepth, a single-stage radar-camera depth estimator trained only with radar, camera, and single-scan LiDAR. All radar returns are aggregated into a fixed-width 1D representation, decoupling runtime from point count. A Height Fusion Block fuses modalities, a lightweight GNN propagates depth globally, and a training-only confidence decoder stabilizes learning with zero test-time cost. We mitigate stripe artifacts via simple augmentations and quantify them using the Vertical-Horizontal Gradient Ratio (VHGR). On nuScenes, compared to recent state-of-the-art methods, JustDepth maintains accuracy while reducing inference time by 39.7x and stripe artifacts by 66% as measured by VHGR.
[CV-25] dRAE: Representation Autoencoder with Hyper-Spherical Codes
链接: https://arxiv.org/abs/2607.22148
作者: Tianren Ma,Lin Long,Chuyan Chen,Mu Zhang,Junbo Zhao,Tong Zhang,Qixiang Ye
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: Preprint. Project Page: this https URL
Abstract:In this work, we aim to discretize the high-dimensional visual representations to bridge the gap with language models - a non-trivial challenge, as existing quantization methods suffer from codebook collapse, failing to scale while preserving semantic coherence. We identify the root cause as metric mismatch: standard Euclidean codebook objectives are fundamentally misaligned with the anisotropic geometry of representation space, leading to codebook embeddings with high-variance magnitude scales and uneven angular distributions that hinder scalability. To address this, we propose Hyper-Spherical Quantization (HSQ), which decouples semantic content from feature magnitude via angular routing, preventing code assignment from being dominated by scale rather than meaning. The resulting discrete Representation Autoencoder (dRAE) achieves high-fidelity reconstruction while preserving semantic integrity and supporting scalable codebook budget. Extensive experiments demonstrate consistent performance gains as the vocabulary size scales to 131,072, along with 100% codebook utilization, simplified training pipeline, and strong performance across understanding and generation tasks.
[CV-26] Visual Relocalization from Sparse Views in Aliased and Low-Texture Environments via Novel View Synthesis IROS2026
链接: https://arxiv.org/abs/2607.22147
作者: Maria Peribañez,Javier Civera,Rudolph Triebel,Riccardo Giubilato
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted to IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2026)
Abstract:Visual localization becomes extremely challenging in planetary-like terrains characterized by low texture, perceptual aliasing, harsh illumination, and sparse, weakly overlapping viewpoints induced by forward rover motion and unconstrained driving directions. Under these conditions, state-of-the-art image-to-image and image-to-map matching pipelines suffer significant performance degradation. In this work, we propose a visual relocalization method that departs from classical correspondence-based pipelines by directly estimating camera poses against a differentiable map representation built with 3D Gaussian Splatting (3DGS). Our key contribution is a geometry-aware training strategy that combines photometric and geometric losses, where the geometric supervision is provided for the first time by combining multi-view stereo (MVS) and LiDAR depths. We show that this joint optimization produces a 3DGS model that better fits the underlying scene geometry, leading to improved photometric and geometric consistency and more robust, accurate single-image 6-DoF pose estimation. Extensive experiments on data acquired in planetary-analog environments validate the effectiveness of our approach, showing substantial gains in relocalization accuracy under challenging conditions. Code is available at this https URL.
[CV-27] CARDIAG: A Dense Segment Classification Benchmark of Deep Learning Architectures for Coronary Angiography
链接: https://arxiv.org/abs/2607.22139
作者: Dominik Bernard Lau,Hubert Malinowski,Jerzy Szyjut,Adam Brzeski,Tomasz Dziubich,Radosław Targoński,Tomasz Figatowski,Natalia Zielińska
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: 19 pages, 12 figures, 3 tables, dataset available on Zenodo at this https URL , code available at this https URL
Abstract:Accurate pixel-level classification of coronary angiograms is critical for cardiovascular disease assessment, yet the field lacks standardized evaluation protocols. In this work we demonstrate a new benchmark for the assessment of deep learning models which densely classify pixels of coronary angiograms to one of SYNTAX classes (or background). The evaluation covers 24 distinct architectures starting with classic convnets to recent state-space-based vision algorithms. We release CARDIAG - a multi-center, multi-label dataset which we carefully split to reliably compute metrics, accounting for diameter error, overlap, centerline quality and calibration. The data contains SYNTAX labels, binary, uncertainty and segmentation masks as well as intermediate frames together with the selected non-sensitive DICOM metadata. From the multitude of algorithms, we nominate ConvNeXt V2 encoder with DeepLab V3 Plus decoder as the best performing, achieving macro F_1=0.456 , which we then ensemble with Mamba U-Net and Feature Pyramid Network, for an increased F_1=0.479 . We demonstrate all the architectures to be well calibrated and determine the generalization of the top 5 methods, together with the data efficiency of these architectures. We highlight the importance of both high-resolution and low-resolution features in encoding. We also demonstrate the model correctness in the context of patient demographic, vessel sides and projection angle configurations. Overall the released benchmark allows for future studies to robustly and rigorously assess the proposals, not only for SYNTAX segmentation, but lesion detection and many more.
[CV-28] GLI-AL: A Multi-Modal Glioma MRI Label Resource with Unified Anatomy-Lesion Labels
链接: https://arxiv.org/abs/2607.22135
作者: Xingyu Xiang,Shuang Hao,Fan Wang,Jianhua Ma,Chunfeng Lian
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Existing BraTS-GLI datasets provide a widely used benchmark for adult glioma MRI segmentation, but their task definition focuses on tumor subregions and does not systematically represent coexisting white matter hyperintensities (WMH). In joint segmentation settings, such unlabeled abnormalities introduce task-specific label noise by treating pathological regions as normal tissue. To address this limitation, we introduce BraTS-GLI Anatomy-Lesion, a controlled-access, labels-only derived resource built from the BraTS 2023-GLI training cohort. The resource provides 1,251 unified eight-class anatomy-lesion label sets aligned with the original four-modal MRI cases, including image-repair labels for 116 cases requiring repaired imaging inputs. The cohort is organized into a 394-case purified subset and an 857-case extended subset, with case-level metadata covering label source, image-repair requirements, quality-control status, access conditions, checksums, and release boundaries. Compared with the original BraTS-GLI annotations, the resource substantially expands foreground supervision by incorporating healthy brain tissues and previously unlabeled coexisting abnormalities within a unified label space. A validation study using MedNeXt and T1/FLAIR inputs suggests that WMH-aware supervision preserves healthy-tissue segmentation performance across both in-domain GLI and external WMH datasets, while improving sensitivity to coexisting lesions relative to noisy-control training. The resource is intended for scientific research and supports joint anatomy-lesion supervision, label-noise analysis, and reproducible evaluation. Data are available at this https URL, and code is available at this https URL. The data resource DOI is this https URL.
[CV-29] A Framework for Individual Tree Growth Reconstruction Using Multi-Platform Laser Scanning
链接: https://arxiv.org/abs/2607.22129
作者: Daniella Tavi,Valtteri Soininen,Lassi Ruoppa,Jesse Muhojoki,Juha Hyyppä
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Accurate tree-level forest monitoring using laser scanning data requires reliable tree delineation, consistent tree correspondence across multitemporal point clouds, and accurate estimation of tree attributes and their change. Reconstructing tree growth in boreal forests is challenging due to the scarcity of historical stem-level data, propagation of errors from older sensors into change estimation, and growth rates with a magnitude of measurement uncertainty. This study investigates a framework for estimating individual tree diameter at breast height (DBH) and stem volume growth using 136 point clouds acquired between 2014–2025 with 11 scanners on airborne (ALS), mobile (MLS), and terrestrial laser scanning (TLS) platforms across boreal forest test sites. Trees were delineated from an MLS point cloud using deep learning-based segmentation which was transferred to the remaining point clouds, resulting in reliable multitemporal tree correspondence. Stem curves were derived from MLS/TLS data, with ALS data used for height estimation, enabling DBH and volume estimation and time series. A height growth-based scaling model was used to reconstruct stem attributes across time and estimate growth. Results showed that modeled growth achieved higher agreement with manual growth estimates than differencing independently estimated attributes from point clouds. The modeled-manual 5- and 10-year growth RMSEs were 55–111% and 26–67% for DBH, and 31–87% and 21–67% for volume, respectively, depending on plot difficulty. The scaling model was temporally robust, with errors remaining stable or stabilizing after 5–6 years, reaching maximum RMSEs of 8–12% for DBH and 12–23% for volume after 12 years. Combining MLS/TLS-derived stem measurements with multitemporal ALS-derived heights provided a robust framework for individual tree growth estimation without requiring multiple under-canopy scans.
[CV-30] Projection Pursuit CPCANet for Domain Generalization
链接: https://arxiv.org/abs/2607.22117
作者: Yu-Hsi Chen,Abd-Krim Seghouane
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 8 pages, 5 tables
Abstract:Domain Generalization (DG) aims to learn representations robust to distribution shifts. Recent geometric alignment methods, such as CPCANet, extract domain-invariant structures through batch-wise Common Principal Component Analysis (CPCA). However, CPCANet suffers from rank-deficient covariance estimation due to the small-sample-size issue in mini-batch training. To address this limitation, we propose Projection Pursuit CPCANet (PP-CPCANet), a covariance-free framework that learns a global orthogonal basis on the Stiefel manifold and jointly optimizes it with network parameters via the Cayley transform. We further introduce a symmetry-breaking detached-median PP dispersion objective to extract common principal components (CPCs) with dense and robust optimization signals. Experiments on four DG benchmarks show that PP-CPCANet achieves SOTA performance while maintaining stable training.
[CV-31] InnoText: A Unified Model for Visual Text Generation and Editing ECCV2026
链接: https://arxiv.org/abs/2607.22101
作者: Haowei Liu,Runze He,Jian Lu,Ao Ma,Run Ling,Ke Cao,Jiasong Feng,Wei Feng,Shuo Lu,Yexing Xu,Yun Wang,Jing Wang,Zhanjie Zhang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted by ECCV 2026
Abstract:Diffusion models have recently achieved remarkable success in high-fidelity image synthesis, yet their application to visual text generation and editing remains relatively underexplored. Unlike general image generation, visual text tasks demand precise structural regularity and legibility, which may pose additional challenges for small-scale text and non-Latin scripts such as Chinese. Existing UNet-based models often struggle to produce clear and coherent text, while DiT-based models, though more expressive, are typically limited to a single task, which may lead to redundant training pipelines, inconsistent visual styles, and reduced cross-task generalization. To address these challenges, we propose InnoText, a unified DiT-based framework capable of performing both text generation and editing within a single model. We introduce a Font Size-Aware Modulation (FSAM) module to enhance representations across font scales, a Small-Character Aware Augmentation strategy to improve fine-grained fidelity, and a Task-Specific Region Weighted Loss for adaptive optimization. To support training and evaluation, we also construct a high-quality bilingual (English-Chinese) visual text dataset covering diverse fonts, sizes, and backgrounds. Experimental results demonstrate that our method achieves superior generation accuracy and editing quality, producing visually appealing and realistic text images.
[CV-32] Spectral Prior for Reducing Exposure Bias in Diffusion Models ECCV2026
链接: https://arxiv.org/abs/2607.22091
作者: Yuya Kobayashi,Masato Ishii,Yuhta Takida,Takashi Shibuya,Yuki Mitsufuji
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted at ECCV2026
Abstract:Diffusion models typically suffer from error accumulation during iterative sampling, commonly referred to as exposure bias. We reveal systematic frequency-dependent discrepancies between training and inference, which can be interpreted as frequency-dependent SNR error. Crucially, the direction of this mismatch varies across models and timesteps, indicating that fixed correction rules do not generalize. We propose Spectral Alignment (SPA), a lightweight, guidance-based method that calibrates the power spectrum of intermediate predictions to a pre-computed prior. Our approach consists of two stages: (1) offline fitting of a parametric spectrum model from training data, and (2) inference-time guidance via efficient FFT-based gradient computation. SPA introduces minimal computational overhead (3-4%) and is complementary to Classifier-Free Guidance (CFG). We demonstrate consistent improvements across diverse architectures, from pixel-space models (DDPM, ADM) to latent diffusion models (SD2.0, SDXL) and flow-matching models (SD3.5, FLUX). Our implementation is available at this https URL.
[CV-33] FAIR: Feature-Augmented Implicit Regularization for AI-generated Fake Image Detection ECCV2026
链接: https://arxiv.org/abs/2607.22087
作者: Md Redwanul Haque,Manzur Murshed,Manoranjan Paul,Tsz-Kwan Lee
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted to ECCV 2026
Abstract:Generalization remains a critical bottleneck in AI-generated image detection. Because many modern generators are proprietary or adversarially modified, existing detectors overfit to the low-level textural patterns of accessible training data, resulting in severe failures on unseen domains. Conventional regularization techniques (e.g., L_1 / L_2 norms, Dropout) apply indiscriminate parametric constraints and fail to provide the domain-invariant structure necessary for cross-generator robustness. To address this, we propose Feature-Augmented Implicit Regularization (FAIR). FAIR introduces an orthogonal, macro-structural prior, specifically, Scene Composition Structure (SCS), during training to geometrically constrain the model’s optimization trajectory. By augmenting the primary feature space with domain-invariant SCS features, FAIR explicitly penalizes texture-biased shortcut learning. Crucially, this structural prior is entirely discarded at inference, yielding a smoothed, generalized decision boundary with zero architectural or computational overhead. Extensive evaluations across five massive benchmarks demonstrate that integrating FAIR into state-of-the-art detectors significantly improves cross-generator generalization, boosting accuracy by up to 8.04% and establishing new state-of-the-art robustness in zero-shot transfer scenarios.
[CV-34] CommandLM: Data driven behavior level descriptor for ego vehicles
链接: https://arxiv.org/abs/2607.22078
作者: Boris Tokic,Constantin Selzer,Fabian B. Flohr
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:As autonomous driving systems move toward real-world deployment, interpretable, behavior-level decision-making is essential for safety, trust, and regulation. We introduce CommandLM, a multimodal large language model that generates concise, human-readable behavior descriptions for ego vehicles from fused multi-sensor data. Our model processes temporally fused bird’s-eye view representations from LiDAR and multi-camera inputs via a Q-Former adapter connected to a quantized, LoRA-fine-tuned large language model. Trained on our CommandLM-nuScenes dataset, CommandLM produces intent-aware, interpretable captions suitable for planner supervision and safety auditing. Experiments demonstrate strong linguistic and behavioral alignment, achieving CIDEr 0.67, and BERT-F1 0.88, substantially outperforming the BLIP-2 baseline (CIDEr 0.52, BERT-F1 0.86). In human evaluation, 58% of the generated descriptions were rated accurate, efficient and rule-compliant, confirming their real-world plausibility. While the remaining descriptions may not always select the most efficient, goal-oriented behavior, CommandLM’s interpretable outputs enable downstream validation systems to identify and correct such cases, making it an effective tool for transparent behavior auditing. These results show that integrating multimodal fusion with language reasoning yields efficient and transparent behavior-level understanding for autonomous driving. We release our code and dataset at: this https URL
[CV-35] Alleviating Regional Shortcuts for Few-Shot Class-Incremental Learning
链接: https://arxiv.org/abs/2607.22072
作者: Haichen Zhou,Yazhe Lyu,Yixiong Zou,Ruixuan Li,Yuhua Li
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted by TMM 2026
Abstract:Few-shot class-incremental learning (FSCIL) aims to incrementally learn novel classes with only a few samples while avoiding forgetting base classes. However, current methods show a tendency to misclassify novel-class samples into base classes, which we find to be caused by the excessive focus on base-class-discriminative regions on novel-class samples. In this work, we aim to explore the underlying mechanism for an interpretation and solution. We first provide a compositional view to analyze the transferred and reused spatial patterns on novel-class samples. Then, through extensive experiments and theoretical analysis, we identify both empirically and theoretically that a shortcut exists in the model’s base-class training, which naturally forms the excessive focus on only the most discriminative regions (primitives), which we term as the regional shortcut. Finally, based on this interpretation, to address this problem, we propose a compositional-learning-based method to learn two primitive sets (a common set and a discriminative set), which alleviates the regional shortcut by constraining the model to learn and utilize the common primitive set for base- and novel-class recognition. Extensive experiments on standard FSCIL benchmarks demonstrate the effectiveness of our approach, yielding consistent improvements over existing state-of-the-art methods in both accuracy and interpretability.
[CV-36] ReCowGnition: A Realistic Biometric Benchmark for Cow Face Recognition ICPR2026
链接: https://arxiv.org/abs/2607.22071
作者: Marco Huber,Marco Kiesewalter,Judith Louise Pieper,Bastian Kubsch,Naser Damer
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted at ICPR 2026 Workshops
Abstract:With the development of precision livestock farming and the advances in computer vision, visual animal biometrics has gained attention. Using biometric technologies that have been proven effective for humans to identify livestock can increase animal welfare as well as production efficiency. However, challenges such as complex scenarios, similar appearances, occlusions, and non-cooperative behavior, as well as the limited amount of publicly available labeled datasets, remain. In this work, we contribute a novel, publicly available cow face benchmark dataset that has been collected in a realistic automatic scenario with 6,838 images of 161 different cows at a dairy farm. In addition to the public dataset, we define two verification and four identification evaluation protocols to foster comparable research in the cow recognition research field. Further, we provide evaluation results on our dataset of six benchmark models, which include models trained on limited data, cross-species fine-tuned models, and zero-shot foundation model approaches.
[CV-37] Rethinking Multi-Branch and Cross-Backbone Fusion for Vehicle Re-Identification in the Foundation-Model Era
链接: https://arxiv.org/abs/2607.22068
作者: Yu Wang,Hongyu Yang
类目: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注:
Abstract:Multi-branch architectures and CNN-Transformer fusion have long been regarded as effective ways to improve vehicle re-identification (Re-ID) by combining complementary representations. In this work, we revisit this assumption in the foundation-model era through a comprehensive empirical study. A single DINOv3-pretrained ConvNeXt trained with a tuned recipe achieves 88.19 mAP on VeRi-Wild Small and 77.47 mAP on VeRi-Wild Large using visual cues alone, matching the strongest protocol-verified metadata-dependent multi-branch baseline. Applying training-free re-ranking further improves performance to 92.38 and 83.68 mAP, respectively. Using this strong baseline together with retrieval-level branch diagnostics, we evaluate whether increasing representational diversity still provides measurable gains. Across both benchmarks, concatenating multiple branches built on a shared backbone changes the best single-branch performance by less than one mAP point while increasing the embedding dimension by 4x, and the resulting representation has an effective rank close to the original feature dimension. We further study cross-backbone fusion using an asymmetric frozen-anchor strategy to combine ConvNeXt and Vision Transformer representations. Despite these favorable conditions, Transformer branches consistently remain 13-15 mAP below the ConvNeXt backbone, and paired per-query bootstrap analysis estimates the largest observed fusion gain to be only +0.11 mAP (95% confidence interval). Our results suggest that, under the evaluated setting, improving a single strong foundation-model backbone together with retrieval-stage re-ranking is more effective than increasing architectural complexity through additional branches or heterogeneous backbones. We restrict our conclusions to single-seed training and one family of foundation models and discuss conditions under which these observations may not hold.
[CV-38] A Smooth Phase-Separation Model for Weak-Boundary Segmentation of Homogeneous Structures
链接: https://arxiv.org/abs/2607.22053
作者: Zihan Li,Jiebao Sun,Fanghui Song,Zhichang Guo
类目: Computer Vision and Pattern Recognition (cs.CV); Mathematical Physics (math-ph)
备注:
Abstract:Segmentation of adjacent structures with similar intensity distributions remains a challenging problem in image analysis, particularly when object boundaries are weak or ambiguous. Under such conditions, classical variational models may suffer from degenerated image-driven forces, leading to boundary leakage or undesired merging of neighboring regions. To address these limitations, we propose a smooth phase-separation variational model based on the Cahn–Hilliard equation for weak-boundary segmentation of homogeneous-appearance structures. The proposed framework integrates softmax-based region fitting with Cahn–Hilliard phase-field regularization to maintain interface discrimination under weak image-driven forces. We further introduce a mixed L^2-H^-1 gradient flow, which preserves higher-order interfacial regularization while allowing adaptive changes of phase masses, establish the continuous energy dissipation law, and prove the existence and uniqueness of weak solutions in the natural solution class. For numerical computation, we develop a stabilized scalar auxiliary variable (SAV) scheme that is linear, FFT-based, and satisfies a modified discrete energy dissipation law. Numerical experiments on synthetic and medical images demonstrate that the proposed method effectively separates adjacent homogeneous structures across weak boundaries and achieves competitive segmentation accuracy and improved boundary localization compared with representative variational, phase-field, and deep learning methods.
[CV-39] EVL-MCoT: Enhanced Vision-Language Multi-CoT for Harmful Meme Detection
链接: https://arxiv.org/abs/2607.22016
作者: Hao Yang,Jin Wang,Xuejie Zhang
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:
Abstract:MEMEs are widely used on the internet and often carry strong elements of sarcasm or irony. Understanding their hidden meanings typically requires a joint interpretation of text and vision. Existing methods focus on the dual-stream vision-language model to extract the visual and text simultaneously, which lacks background information and prior knowledge about the comprehensive explanation of MEME. One feasible option is to adopt chain-of-thought (CoT). However, the simple CoT approach lacks multi-perspective thinking, which may compromise the reliability of the resulting answers. Moreover, it often relies on shallow feature fusion, lacking the fusion of local details and fine-grained visual-prompt text alignment. This limitation prevents a deeper understanding of the intricate connections between the visual and the text. Herein, an enhanced vision-language multi-CoT (EVL-MCoT) approach is proposed to address these limitations. By promoting multi-CoT, EVL-MCoT enhances consistency and reduces bias in the decision-making process. Additionally, we design a prototype-guided and context-guided decoding framework, which incorporates visual prototypes to guide the fusion process and enables the model to align textual and visual information more precisely. We achieve promising results on the HatefulMemes and MultiOff datasets. The source code has been publicly released and is available at this https URL.
[CV-40] Visual Saliency Steering Distillation for Multimodal Chain-of-Thought Reasoning
链接: https://arxiv.org/abs/2607.22013
作者: Hao Yang,Jin Wang,Xuejie Zhang
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:
Abstract:Multimodal chain-of-thought (CoT) reasoning integrates visual and textual cues through step-by-step inference. In small models with limited token budgets, modality-interaction fusion often suppresses tiny cross-modal differences. In particular, multimodal CoT often struggles when different images pair with identical text or different texts pair with an identical image, making such inputs nearly indistinguishable after fusion. This study proposes Visual Saliency Steering Distillation (VSSD). VSSD leverages the attention maps of multimodal large language models to generate perturbed images that capture task-sensitive feature directions, and then applies singular value decomposition to extract dominant steering vectors to guide inter-layer distillation. Experiments on ScienceQA and M ^3 CoT demonstrate that VSSD improves rationale generation and answer inference. The code is available at this https URL.
[CV-41] Medical-Checklist: Assessing the Comprehension of Medical Images by Multimodal Models ALT
链接: https://arxiv.org/abs/2607.21998
作者: Bannapol Limanond,Masanori Suganuma,Takayuki Okatani
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted for publication in IEEE Journal of Biomedical and Health Informatics
Abstract:This paper introduces a new benchmark test, Medical-Checklist, for assessing medical multimodal models. The recent advancements in multimodal models have demonstrated significant potential in the field of medical vision-language tasks. However, it is becoming increasingly clear that evaluating these models’ performance, whether they are applied to natural or medical images, is challenging. The critical question is whether the models can accurately understand an input image while associating it with relevant input text. To address this, Medical-Checklist imposes a binary test on the models: they are given an image and two captions, where one is correct and the other incorrect, and the model must select the correct one. The incorrect caption contains a single medical concept (word or phrase) that is inaccurately substituted from the correct caption. Although the task is simple, this simplicity enables the unified assessment of diverse multimodal models designed and learned on different principles. It also enables us to verify whether models correctly understand a wide range of medical concepts across various medical sub-domains. Medical-Checklist is designed to reduce potential biases in data and to enable evaluation of the models’ ability to handle out-of-distribution inputs, which were difficult in existing datasets. When evaluating four state-of-the-art medical multimodal models with Medical-Checklist, it was revealed that despite their excellent performance in specific tasks such as Med-VQA, they may not correctly understand images, suggesting a long journey ahead for clinical application. The dataset and code will be made public upon acceptance.
[CV-42] Rethinking Layer-Wise Information Allocation for Vision Foundation Model Adaptation ACM-MM2026
链接: https://arxiv.org/abs/2607.21973
作者: Yuqi Li,Xi Xiao,Yunbei Zhang,Lin Zhao,Yu Li,Aiden Zhao,Tianyang Wang,Hao Xu,Yingli Tian
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted by ACM MM 2026
Abstract:Vision foundation models are increasingly reused as frozen backbones for downstream visual recognition, making parameter-efficient adaptation a central problem. Prompt-based adaptation, including Visual Prompt Tuning (VPT), provides a lightweight way to specialize these models, but its layer-wise behavior remains poorly understood: performance is sensitive to prompt depth, placement, and task distribution, and gains on standard in-domain benchmarks do not always translate into robust generalization. We argue that this limitation is not solely an optimization issue, but a layer-wise information allocation issue: existing prompt-based methods lack principled control over what prompt-conditioned representations should preserve, suppress, and propagate across depth. Inspired by the Information Bottleneck principle, we introduce Prompted Information Bottlenecks (PIB), a framework that regularizes layer-wise compression-sufficiency trade-offs and promotes a more coherent cross-layer information path. The key idea is that effective adaptation should be minimal yet sufficient, retaining task-relevant local evidence in earlier layers while progressively discarding nuisance factors and redundant details in deeper layers. Extensive experiments show that PIB achieves strong performance across 34 datasets, reaching 92.1% on FGVC, 93.01% on HTA, and 77.33% on VTAB-1k, while tuning only 0.35% parameters on average across the main settings. Beyond benchmark accuracy, PIB helps explain the non-monotonic behavior of prompt capacity scaling, reduces shortcut reliance, and improves robustness under distribution shift and fine-grained recognition settings. These results position PIB as both a practical method and an information-allocation perspective for adapting frozen vision foundation models. Our code is available at this https URL
[CV-43] xtSLIP: Text Self-Supervised CLIP for Medical Report Generation
链接: https://arxiv.org/abs/2607.21970
作者: Haoyu Jiang,Ziping Cong
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: 9 pages
Abstract:Automating radiology report generation is important for improving reporting consistency and clinical workflows . While Contrastive Language–Image Pretraining (CLIP) has advanced medical vision language modeling, existing CLIP-style approaches may still provide insufficient fine-grained semantic supervision for complex report generation. Standard CLIP primarily optimizes cross-modal alignment, without explicitly structuring the textual embedding space that guides visual representation learning. To address this limitation, we propose TextSLIP, a general medical vision-language pretraining framework that augments CLIP with intra-modal text contrastive learning. By improving textual embedding discriminability through self-supervised augmented text pairs, TextSLIP is designed to provide finer-grained linguistic supervision to the visual encoder. As an initial validation, we pretrain TextSLIP on a curated dataset of 7 million brain MRI image-text pairs and fine-tune the pretrained visual encoder within a report generation architecture. In controlled comparisons with CLIP-style baselines, TextSLIP shows consistent improvements on report generation metrics. Ablation studies further suggest that text-side self-supervision contributes to the observed gains. These results indicate that text-level contrastive learning is a promising direction for improving medical visual-textual alignment, while broader validation across additional medical domains remains an important next step.
[CV-44] Low-Altitude Channel Multipath Prediction via Panoramic Perception and Vision-Language Model
链接: https://arxiv.org/abs/2607.21953
作者: Zihang Zeng,Shu Sun,Meixia Tao,Zhiyong Chen,Jianhua Mo,Xiangwen Gu
类目: Computer Vision and Pattern Recognition (cs.CV); Signal Processing (eess.SP)
备注:
Abstract:Unmanned aerial vehicle (UAV) communication is expected to support a wide range of low-altitude applications in 6G mobile networks. However, traditional statistical channel models provide limited accuracy in specific environments, while deterministic methods such as ray tracing usually rely on accurate three-dimensional environment models and involve high computational complexity. Existing multimodal channel prediction approaches mainly focus on large-scale metrics such as path loss, and remain insufficient for modeling small-scale parameters. To address these limitations, this paper proposes PanoLAMP, a Panoramic perception and vision-language model-based Low-Altitude Multipath Prediction framework. It adopts a pretrained vision-language model as the backbone and captures the propagation environment features through panoramic RGB-D observations collected at both the transmitter and receiver to predict the delay, power, azimuth angle, and zenith angle offset relative to the line-of-sight path. Experiments are conducted on a synthetic dataset containing 18,949 UAV-vehicle links across seven UAV altitudes. Experimental results show that the proposed method consistently outperforms representative baselines in both multipath parameters and statistical metrics, and demonstrates stronger generalization across different flight heights.
[CV-45] ISPCloak: Weaponizing ISP for Optimization-Free Physical Camouflage against Deepfake Detectors ACM-MM2026
链接: https://arxiv.org/abs/2607.21897
作者: Jiale Zhao,Jiajun Wan,Lei Tang,Ye Qin,Kebing Jin,Jinghui Qin
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI); Cryptography and Security (cs.CR)
备注: Accpted by ACM MM 2026
Abstract:The rapid advancement of generative models has spurred the critical need to evaluate the worst-case robustness of deepfake detectors. In this paper, we reveal a fundamental blind spot in current forensic paradigms: while existing detectors excel at capturing digital synthesis artifacts, their effectiveness drops drastically when AI-generated content is cloaked in authentic physical imaging characteristics. We posit that genuine photographs inherently possess hardware-intrinsic statistical signatures, which are imperceptible footprints imprinted by optical sensors and Image Signal Processing (ISP) pipelines, and are fundamentally absent in purely data-driven generative models. Driven by this insight, we propose ISPCloak, a novel optimization-free adversarial attack framework that explicitly weaponizes the ISP pipeline to mislead the judgment of deepfake detectors. Rather than relying on computationally expensive gradient perturbations, our method first employs an Invertible ISP network to project images into the RAW domain. Then, we seamlessly imprint the complex statistical priors of real cameras onto AI-generated images by injecting realistic Poisson-Gaussian sensor noise and conducting forward ISP reconstruction. Synergized with generative artifact suppression and adaptive masking, this streamlined physical simulation enables ultra-fast generation of adversarial examples. Extensive experiments show that embedding authentic physical perturbations fundamentally disrupts a broad range of current detection mechanisms, yielding universally evasive adversarial examples with imperceptible visual alterations.
[CV-46] Learning Adaptive Semantic Gaussian Allocation for 3D Occupancy
链接: https://arxiv.org/abs/2607.21896
作者: Kanglin Ning,Yiran Zhao,Wenrui Li,Houde Quan,Qifan Li,Xingtao Wang,Xiaopeng Fan
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Semantic 3D Gaussians provide a compact representation for 3D semantic occupancy prediction by rendering semantic primitives into a voxel volume under voxel-wise supervision. Recent methods have improved the modeling ability and efficiency of this representation through more flexible primitive shapes, geometry-guided initialization, and progressive densification. However, these advances mainly determine how primitives are represented, initialized, or added, and do not explicitly address how to select the most useful Gaussians when their total number must be limited to control memory and computation. This imbalance creates an allocation bottleneck: redundant Gaussians remain in simple regions, while difficult regions receive insufficient semantic support. We propose the Semantic Gaussian Allocation Transformer (SAGFormer), which uses Gaussian attributes and local geometric-semantic features to score candidates and select a fixed final Gaussian set. Experiments on nuScenes-SurroundOcc and SSCBench-KITTI-360 show that SAGFormer improves occupancy prediction under the evaluated protocols and yields more semantically consistent and better-utilized Gaussian representations. Under similar final counts and raw coverage, it reduces semantic mixing, strengthens class-consistent voxel support, and produces fewer unused Gaussians. The results indicate that explicit capacity allocation is a useful complement to Gaussian refinement for semantic occupancy prediction.
[CV-47] Farmland Extent and Visible Boundary Mapping from 1 m NAIP Imagery Using Residual U-Net and Text-Prompted SAM 3 Refinement
链接: https://arxiv.org/abs/2607.21881
作者: Mohammadreza Narimani,Vikram Anand,Parastoo Farajpoor
类目: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG); Image and Video Processing (eess.IV)
备注: 12 pages, 7 figures, 3 tables. Preprint submitted to Taylor Francis. Code: this https URL Dataset: this https URL
Abstract:Agricultural field maps are often proprietary, incomplete, or outdated, yet they provide the spatial framework for crop monitoring, production accounting, and land-conversion analysis. This study presents a reproducible workflow for mapping farmland extent and visible boundaries from 1 m NAIP RGB imagery. Thirty-seven scenes spanning open cropland, peri-urban interfaces, semi-arid irrigation geometries, and fragmented mosaics were annotated in CVAT and converted to binary masks. Non-overlapping 256 x 256 patches yielded 5,698 samples, split by source scene into 3,850 training, 770 validation, and 1,078 test patches. A residual U-Net (ResUNet) trained with a Dice-dominant loss, L = 2.5(1 - Dice) + BCE, achieved test accuracy 0.8808, IoU 0.8605, Dice 0.9234, precision 0.8766, and recall 0.9794. A frozen SAM 3 branch prompted with “agricultural farmland field” was fused with ResUNet by logical OR. On selected difficult patches, Dice improved from 0.858 to 0.955 (orchard rows) and from 0.804 to 0.903 (fragmented parcels). Sliding-window stitching produced coherent regional masks (example tile Dice 0.898 and 0.919). The product is a semantic farmland-extent layer, not a cadastral parcel map, and supports agricultural monitoring where current field layers are unavailable.
[CV-48] SCALE: Self-Supervised Constraint-Aware Layout GEneration for Local PR DRV Fixing at Advanced Nodes
链接: https://arxiv.org/abs/2607.21850
作者: Chia-Tung Ho,Haoyu Yang,Guanglei Zhou,Yoshi Nishi,Yaguang Li,Walker Turner,Cunxi Yu,Yiran Chen,Brucek Khailany
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: 8 pages, 5 Figures, 6 Tables
Abstract:As semiconductor manufacturing advances toward sub-2nm nodes, local place-and-route (PR) design-rule violation (DRV) fixing is increasingly limited by complex rule interactions, dense multi-layer routing geometries, and foundry-specific constraints. While Large Language Models (LLMs) have recently demonstrated strong capabilities in EDA scripting and documentation, their application to visual layout understanding remains largely unexplored: diagnosing DRC violations from layout imagery demands precise geometric reasoning and foundry-specific rule knowledge absent from general-purpose VLM training. We propose SCALE, a framework with a self-supervised layout-generation stage for local DRV fixing at advanced nodes. Multi-layer layout geometry is serialized into structured text, and a fine-tuned language model learns to reconstruct randomly masked polygons from surrounding BEOL context alone without violation labels. At inference, natural-language rule constraints and high-temperature sampling steer generation toward diverse, violation-prone layout variants validated by an industrial signoff DRC checker, producing DRC-annotated layout–violation pairs used to fine-tune a domain-adapted DRC-VLM. This VLM provides rule-aware geometric guidance for local DRV repair, boosting state-of-the-art agents’ solve rates by +12–25% (up to 97%) on 100 real sub-2nm cases spanning enclosure, spacing, width, and color-spacing violations.
[CV-49] Closing the Loop: Training-Free Revisit Consistency for Autoregressive Generative Rendering
链接: https://arxiv.org/abs/2607.21848
作者: Wenchao Ma,Changran Liu,Sharon X. Huang,Haomiao Jiang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Recent conditional video generation models have shown promising potentials to transform 3D engine renderings, such as depth maps and untextured geometry, into photorealistic videos for gaming and immersive content creation. These applications require long-horizon auto-regressive generation that continuously synthesizes new frames while preserving a persistent 3D world. Auto-regressive generators synthesize video chunk by chunk with a bounded KV cache, so when the camera revisits a location after its context has been evicted, the model often regenerates inconsistent appearance, even though the conditioning renderings (e.g., depth) remain perfectly aligned with the underlying this http URL address this revisit inconsistency without any post-training by exploiting correspondences the 3D engine already provides: temporal correspondence retrieves pose-matched historical latent chunks into the KV cache as loop-closure memory, while spatial correspondence from camera pose and depth reprojection biases token-level attention toward geometrically corresponding regions of the retrieved chunks. We demonstrate our method on loop-closure trajectories mined from TartanAir and TartanGround dataset to mirror complicate real-world application scenarios, where it outperforms existing training-free baselines on revisit consistency without losing overall video quality. Project Page: this https URL
[CV-50] oward High-Fidelity 3D Point-Cloud Learning for Brain Folding Morphology Prediction Using Trans-Unet
链接: https://arxiv.org/abs/2607.21840
作者: Geran Zhao,Xiaotian Li,Poorya Chavoshnejad,Mir Jalil Razavi,Akbar Solhtalab,Lijun Yin,Guifang Fu
类目: Computer Vision and Pattern Recognition (cs.CV); Image and Video Processing (eess.IV); Methodology (stat.ME); Machine Learning (stat.ML)
备注:
Abstract:Learning high-fidelity point-cloud features in the 3D space poses significant challenges, including permutation invariance, lack of local context, difficulty in fine-grained surface reconstruction, and high computational cost. In this article, we propose Trans-Unet, a novel framework that addresses these issues by first tansforming 3D point-cloud data into a 2D grid domain and then employing a U-shaped hybrid model that integrates Convolutional Neural Networks, and self-attention mechanisms. The proposed Trans-Unet effectively learns and reconstructs precise features from high-resolution 3D point-cloud data (with 40,401 points in surface and 2,382 points in fiber) derived from a predefined finite element brain patch growth model, enabling accurate prediction of brain folding patterns. By combining multiple techniques, Trans-Unet leverages the complementary strengths: the 3D-to-2D transformation preserves fine-grained structural information while significantly reducing computational cost and the curse of dimensionality; convolutional blocks capture hierarchical, low-level local representations; and the self-attention mechanism models global, high-level semantics and long-range dependencies. The dataset consists of 3D point-clouds containing both brain surface patches and fiber information generated by a large-scale finite element model. Trans-Unet is applied to predict brain surface folding from the initial state (state 0 or states 0-2) to the final state (state 3). Experimental results demonstrate that Trans-Unet achieves high-resolution predictions of brain patch growth, surpassing existing methods in both fidelity and accuracy.
[CV-51] What Happens to Accuracy When Photo Lineups Contain Non-Mated Rank-One Images From Large Galleries?
链接: https://arxiv.org/abs/2607.21792
作者: Genesis Argueta,Kevin W. Bowyer,Michael King,Jayeeta Dhar
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 22 pages, 18 figures, the last 8 pages are figures of suspects used
Abstract:One-to-many facial identification is commonly used to match a probe image from surveillance video against a gallery of driver’s licenses and/or booking photos. The algorithm’s rank-one image from the gallery, or a human examiner’s selection from the algorithm’s top-ranked images, may then be placed in a photo lineup shown to a witness. Witness selection of the gallery image in the photo lineup may then lead directly to the person in the gallery image being arrested. This facial identification process is involved in at least 9 wrongful arrests. This work specifically examines whether the probability of a witness making an incorrect identification increases with the size of the gallery searched. We compare photo lineup accuracy when the “suspect” image is drawn from galleries of 500, 5,000, and 24,000 images. We find that larger galleries increase both the likelihood of a witness making an incorrect identification and their confidence in that (incorrect) identification. These results raise questions of whether an image resulting from such a facial identification process should be used in photo lineups and of whether results of a photo lineup alone should constitute probable cause for arrest.
[CV-52] Risk-Routed Implicit Boundary Refinement for Robust Ultrasound Image Segmentation
链接: https://arxiv.org/abs/2607.21787
作者: Jingguo Qu,Xinyang Han,Xiang Wang,Yuqi Yang,Tonghuan Xiao,Sheng Ning,Jing Qin,Ann Dorothy King,Winnie Chiu-Wing Chu,Jing Cai,Michael Ying
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Medical ultrasound (US) image segmentation faces significant challenges due to speckle noise, low-contrast boundaries, acoustic shadowing, and acquisition variation across operators and clinical centers. Although encoder-decoder and transformer-based networks have achieved strong performance, many methods recover boundary details through dense decoders or larger backbones, which may still produce over-smoothed contours or unstable predictions under external distribution shifts. In this article, we propose Risk-routed Implicit Boundary Refinement (RIBR), a compact segmentation framework that uses implicit neural representation as a risk-routed residual correction rather than an unconstrained full-mask predictor. RIBR combines boundary-refinement implicit residuals, risk-routed residual control, and geometry- and speckle-aware boundary regularization to refine uncertain contours while suppressing non-boundary oscillations. Evaluation on nine US datasets covering lymph nodes, breast lesions, thyroid nodules, and prostate shows that RIBR achieves the best overall macro-average and consistently reduces boundary error across grouped and organ-specific comparisons under a compact parameter budget. These findings suggest that controlled implicit residual learning is a practical strategy for resource-constrained and boundary-sensitive US segmentation. Source code is available at this https URL.
[CV-53] Physiological Signals as a Forensic Modality for Talking-Face Deepfake Detection
链接: https://arxiv.org/abs/2607.21776
作者: Othmane Harraq,Tamer Aldwairi
类目: Machine Learning (cs.LG); Cryptography and Security (cs.CR); Computer Vision and Pattern Recognition (cs.CV); Multimedia (cs.MM)
备注:
Abstract:Talking-face (TF) deepfake generation synthesizes photore- alistic facial video from a static source image and an au- dio signal, producing forgeries that current image-based detectors consistently fail to identify. Unlike face-swap ma- nipulation, TF synthesis has no underlying real video from which to inherit physiological characteristics, making re- mote photoplethysmography (rPPG) a uniquely motivated detection modality for this forgery category. We propose a detection framework that extracts per-video rPPG wave- forms via RhythmFormer and trains a suite of lightweight classifiers to distinguish real from synthesized physiologi- cal signals. Evaluated on the TF subset of Celeb-DF++ un- der a strict subject-independent protocol, where test identi- ties are completely separated from training identities, our 1D ResNet achieves an AUC of 0.806 and EER of 27.8%, placing it within 2.4 points of the best published general- purpose detector (Effort, ICML 2025) while operating ex- clusively on the physiological channel. We document a con- trolled reproduction study of DeepFakesON-Phys, the rep- resentative prior rPPG detector, demonstrating degrada- tion from AUC 0.999 on legacy face-swap data to 0.622 on the TF subset of Celeb-DF++. We further show that detec- tion difficulty is strongly method-dependent: AUC ranges from 0.985 (Real3DPortrait) to 0.690 (IP-LAP) across the seven TF generators, with the ranking remaining perfectly stable across all evaluation protocols. This spread reflects an interpretable physiological property of each generator rather than evaluation noise, and constitutes the primary theoretical contribution of the work.
[CV-54] Be Consistent! Enhancing Robust Visual Reasoning in LVLMs with Consistency Constraints
链接: https://arxiv.org/abs/2607.21722
作者: Liqiang Jing,Xiong Zhou,Siddharth Varia,Neha Anna John,Xinya Du,Vassilis N. Ioannidis
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:While Large Vision-Language Models (LVLMs) exhibit strong perceptual capabilities, they remain vulnerable in visual reasoning tasks. Existing benchmarks largely focus on symbolic mathematical or scientific problems and simple vision-centric tasks, offering limited assessment of complex visual reasoning and logical consistency, a critical requirement for reliable reasoning systems. We introduce ConVBench, a complex vision-centric reasoning benchmark in which each image is paired with two logically equivalent questions across six categories: action and state, complex counting, spatial reasoning, causal and intent understanding, commonsense reasoning, and temporal perception. To complement this benchmark, we define two evaluation metrics, logical consistency and robust accuracy, that jointly assess both the correctness and consistency of model responses. We further present ConVLM, which improves LVLM reasoning through Group Relative Policy Optimization (GRPO)-based reinforcement learning with a novel consistency reward. This method leverages automatically generated logically equivalent question-answer pairs and a dual-reward design combining accuracy- and consistency-based signals, encouraging agreement between paired responses. The framework functions effectively with or without strict answer supervision.
[CV-55] Oxygen-TryOn: Fashion-Native Foundation Model for Any-item Virtual Try-On
链接: https://arxiv.org/abs/2607.21694
作者: Yong Liu,Xiaolong Fu,Zihang Xu,Wen Xue,Xueheng Li,Lin Song,Yuan Zhang,Chuyang Zhao,Haoyang Huang,Nan Duan,Yipeng Sun,Yan Li,Simiu Gu
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:We present Oxygen-TryOn, a unified foundation model for any-item virtual try-on. Rather than repurposing a general-purpose image editor, Oxygen-TryOn is fashion-native, built for try-on through a dedicated data engine and try-on-specific training. Given one or more reference items (clean product shots or in-the-wild worn-on photos) and a single target subject image, it synthesizes a photorealistic image of the subject wearing the items across virtually any fashion category. Prior systems handle a single garment category in a studio setting, and recent multi-reference methods remain garment-centric; in contrast, Oxygen-TryOn supports diverse items and scenarios, including full- and half-body views, a variable number of references, and free multi-item composition, while faithfully preserving both subject identity and item appearance. Instead of mask-based inpainting, we reformulate try-on as a multi-reference, understanding-driven generation task. We build a data engine that collects, manufactures, annotates, and filters high-quality try-on data at scale, and design a three-stage recipe of continued pre-training (CPT), supervised fine-tuning (SFT), and reinforcement learning (RL). The RL stage uses a hybrid reward combining an in-house try-on reward model with a proprietary, rubric-guided general-purpose model, jointly supervising fine-grained consistency and instruction-level quality. It also follows general editing instructions (e.g., pose changes) in the same pass. Across public benchmarks and our in-house Oxygen-TryOn Bench, it achieves state-of-the-art consistency and realism on single-item try-on and leads on multi-item try-on, matching or surpassing both leading proprietary systems (Nano Banana Pro, GPT-Image-2, Seedream5 Lite) and open-source models (FLUX.2).
[CV-56] Self-Poisoning in Adaptive Out-of-Distribution Detection: A Sharp-Threshold Theory and Certified Label-Free Calibration
链接: https://arxiv.org/abs/2607.21673
作者: Vishnu Bindu Balachandran
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Cryptography and Security (cs.CR); Computer Vision and Pattern Recognition (cs.CV); Machine Learning (stat.ML)
备注: 28 Pages, 5 figures
Abstract:Test-time adaptive out-of-distribution (OOD) detectors update a memory bank from the unlabelled stream. We show this adaptation obeys a provable dynamical law. Modelling bank impurity as a generalized Pólya urn, we prove almost-sure convergence to a mean-field equilibrium whose slope acts as a reproduction number. Below one, impurity stays benign. Above one, the bank is fully poisoned and the detector collapses. The measured admission kernel is affine ( R^2 \ge 0.996 ) with slope just below one in every encoder family (a protocol signature), so this detector class is near-critical by design, and across 96 settings the predicted threshold matches the empirical collapse, where ungated dictionaries lose up to 0.163 AUROC. We then prove that a certified admission gate, reading only a frozen reserve, severs the feedback loop and removes the transition at every contamination rate, even adversarially, while controlling false positives label-free. For the complementary static-calibration failure under drift we give CDC, which restores nominal FPR label-free on all tested drift-affected cells. Finally we prove a two-world impossibility theorem. Drift and contamination are indistinguishable without labels, forcing a closed-form power ceiling our procedure approaches. Together these give a complete possibility/impossibility characterization of label-free adaptive OOD detection.
[CV-57] Pixels for Programs? A Cross-Provider Case Study of Input-Token Accounting for Source Code as Text and Images
链接: https://arxiv.org/abs/2607.21672
作者: Ronak Bhalgami
类目: oftware Engineering (cs.SE); Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Long source-code contexts consume many text tokens, motivating the proposal to render code as images for vision-language models. Recent work asks whether models can still solve code tasks after this transformation. We examine a different systems question: how commercial APIs count the resulting requests. We present a reproducible measurement case study of provider-reported input tokens for raw source text and a compact rendered-image representation. The benchmark pairs requests across five programming languages, nine source lengths from 20 to 2,000 lines, and 15 available model aliases exposed by Anthropic, OpenAI, and Google Vertex AI. These aliases collapse to approximately five distinct accounting signatures and are not independent model replications. Across 675 complete text/image pairs, aggregate image-to-text ratios are 0.135, 0.194, and 0.242, corresponding to reported input-token reductions of 86.5%, 80.6%, and 75.8%, respectively. These totals conceal materially different break-even behavior: Anthropic and OpenAI images receive lower counts at every tested size, while Gemini images require 6.95 times as many tokens at 20 lines and cross below text only at 200 lines in the aggregate. A targeted audit also reproduces non-monotonic Gemini image accounting across a page boundary. This study measures black-box request accounting for one compact rendering pipeline. It does not measure semantic fidelity, task accuracy, latency, monetary cost, or coding-agent efficiency. We release the scripts, revision-pinned corpus specification, raw usage records, validators, and deterministic analysis needed to reproduce and extend the study.
[CV-58] Local Synaptic Rules Can Implement a SIGReg Gradient Without Backpropagation
链接: https://arxiv.org/abs/2607.21622
作者: Martin Andrews
类目: Neural and Evolutionary Computing (cs.NE); Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注: 10 pages, 1 figure. Associated code at this https URL
Abstract:We prove that two canonical local synaptic learning rules, the potentiation arm of spike-timing-dependent plasticity (STDP ^+ ) and homeostatic plasticity (instantiated here via flashlight granule-cell-like neurons), together can implement the exact gradient of a SIGReg-like self-supervised learning objective. The equivalence requires no gradient calculations, no global error signals, no weight transport, and no label information: the only inputs are pre- and post-synaptic firing rates, local firing statistics, and the temporal contiguity of natural sensory streams. On a synthetic clustering task designed to probe whether class structure can be recovered from temporal ordering of inputs alone, ordered presentation raised cluster separation (CSR) to 2.49 while random ordering left it near baseline (0.83), a roughly threefold ( \approx 3.5\sigma ) separation attributable solely to input ordering. On temporally ordered MNIST, a two-layer network trained entirely with these rules achieved 87.3% linear-probe accuracy, showing that the mechanism functions end-to-end.
[CV-59] An Explainable FFT-Based Spatial-Frequency Fusion Framework for Deepfake Detection
链接: https://arxiv.org/abs/2607.17441
作者: Pamela Kirui,Cho Hyuk,Qingzhong Liu,Haodi Jiang
类目: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注:
Abstract:Deepfake generation has raised growing concerns regarding digital media authenticity, misinformation, identity fraud, and public trust. Recent studies show that combining spatial and frequency features leads to stronger detection results than using independently. This paper presents MSCA-FFT, a Fast Fourier Transform (FFT)-based multi-scale cross-attention framework for image-level deepfake detection. The model combines a partially fine-tuned Xception spatial branch with an FFT-based frequency branch. The frequency branch processes the log-scaled FFT magnitude spectrum through shallow convolutional layers, avoiding inverse frequency-to-image reconstruction used in DCT-based pipelines. The spatial and frequency representations are refined by transformer encoders, fused through cross-attention, and passed to an MLP classifier for real/fake prediction. Experimental results show that MSCA-FFT achieves consistently higher performance than the DCT-based state-of-the-art spatial-frequency fusion method and the compared baseline models. The ablation study further indicates that the FFT-based frequency branch provides complementary spectral cues when fused with spatial features. In addition, FFT-based frequency analysis and Grad-CAM/LIME explanations show consistent evidence around manipulation-sensitive facial regions, including the eyes, mouth, nose, and facial boundaries.
[CV-60] he Lift Spectrum: How Measurement-to-Space Adaptivity Shapes Robustness in Image-Free Single-Pixel Sensing
链接: https://arxiv.org/abs/2607.22077
作者: Yuyuan Han,Jingwei Li,Long Qiu,Chong Wang,Wenxuan Hao,Jiangyu Han,Xinyu Yao,Yuchen He,Hui Chen,Jianbin Liu,Huaibin Zheng
类目: Image and Video Processing (eess.IV); Computer Vision and Pattern Recognition (cs.CV); Optics (physics.optics)
备注: 26 pages (14 main text + 12 supplementary material), 9 figures. Submitted to IEEE Transactions on Computational Imaging
Abstract:Single-pixel sensing encodes a scene as a short sequence of coded measurements, and image-free methods infer the task directly from that sequence. Removing reconstruction does not remove the difficulty: it relocates it to the lift, the map from 1D measurements to a 2D representation, which prior work treats as a trivial reshape. We recast the lift as the central design axis of image-free sensing and order methods by how strongly it adapts to its input: a fixed-physics inverse (reconstruct-then-segment), a learned static projection, or a content-adaptive retrieval; position on this lift spectrum predicts behavior as acquisition degrades. The spatiotemporal soft-fusion (STSF) network pairs a probe-selected recurrent encoder with a cross-attention lift chosen by a parameter-matched ablation, ahead of its U-Net++ decoder, and trains under task-prioritized loss scheduling (TPLS), a scheduled reconstruction prior. In simulation, STSF+TPLS surpasses the prior image-free baseline on three datasets at 3.13% sampling (+3.2 to +9.9 pp foreground mIoU) and plateaus down to 0.39%. The strongest clean-trained reconstruct-then-segment baseline wins the noiseless limit, but under calibrated measurement noise image-free inference overtakes it, for a measured reason: the reconstruction pipeline amplifies the identical measurement noise before its segmenter reads it. Each region fails in its own signature: collapse, imprinting, or coarsening. STSF+TPLS transfers without fine-tuning to a real single-pixel bench as a proof of concept, at about 14 ms per mask. Charting the lift turns a scattered design space into a map of which lift to deploy at each operating point. Code and pretrained weights: this https URL
[CV-61] A Dual Path Framework with Hotspot Guided Fusion for Three Dimensional CT to PET Synthesis in Head and Neck Cancer
链接: https://arxiv.org/abs/2607.21800
作者: Mohd Maaz Khan,Oluwaseyi Oderinde
类目: Medical Physics (physics.med-ph); Computer Vision and Pattern Recognition (cs.CV)
备注: 25 pages, 11 figures
Abstract:18F-FDG PET/CT plays a central role in staging, treatment planning, and response assessment for head and neck cancer by providing functional information that complements anatomical CT imaging. However, PET acquisition requires radiotracer administration, specialized infrastructure, and additional cost, limiting its availability for repeated imaging. We present a proof of concept deep learning framework for synthesizing PET like images directly from routine CT scans with the goal of providing complementary metabolic information that may support imaging triage and clinical decision support rather than replace diagnostic PET. Forty-four patients from the publicly available QIN-HEADNECK dataset were retrospectively analyzed using five fold cross-validation. We propose a fully three dimensional dual path architecture consisting of (i) a regression U-Net optimized for voxel-wise quantitative SUV estimation and (ii) a conditional generative adversarial network optimized for realistic PET texture. Their outputs are integrated using hotspot guided Laplacian pyramid blending, allowing quantitative information from the regression pathway to be preserved within metabolically active regions while leveraging adversarial texture synthesis elsewhere. The proposed framework achieved a mean absolute error of 0.00395, PSNR of 39.19 dB, and SSIM of 0.9634 on reconstructed three dimensional PET volumes. Qualitative evaluation demonstrated accurate localization of many FDG-avid lesions while producing anatomically realistic background texture. Consistent with previous CT to PET synthesis studies, the principal limitation was systematic underestimation of SUV within highly metabolically active tumor regions.
[CV-62] Hash-QNeRF: Multiresolution Hash Encoding for Quantum Neural Radiance Fields
链接: https://arxiv.org/abs/2607.21675
作者: Digonto Biswas,Tana Ballove,Anjan Bandyopadhyay,Sutanu Mangal,Arun Kumar Pati
类目: Quantum Physics (quant-ph); Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Neural Radiance Fields (NeRF) have revolutionized novel view synthesis, yet their classical implementations remain computationally intensive for high-fidelity rendering. QNeRF recently demonstrated the feasibility of training NeRF on gate-based quantum computers by combining amplitude embedding, parameterized quantum circuits (PQCs), parity-based measurements, and volumetric rendering. However, QNeRF relies on classical sinusoidal positional encoding for spatial coordinates, which scales poorly with scene complexity and resolution. In this work, we replace the sinusoidal positional encoding for spatial coordinates with the multiresolution hash encoding from Instant-NGP while keeping the view-direction encoding, amplitude MLP, quantum circuit, parity measurement, output scaling, and volumetric rendering pipeline unchanged. This hybrid design, Hash-QNeRF, retains the quantum radiance prediction step while benefiting from the fast convergence and memory efficiency of learnable hash grids. On a synthetic Blender scene, we achieve a final training loss of 0.003534, corresponding to approximately 24.5 dB PSNR on the fitted batch. Noise resilience experiments using Qiskit FakeKyiv and FakeTorino backends yield state fidelities of 0.93 to 0.98, indicating that hash encoding does not degrade the quantum circuit’s noise tolerance. Subjects: Quantum Physics (quant-ph); Computer Vision and Pattern Recognition (cs.CV) Cite as: arXiv:2607.21675 [quant-ph] (or arXiv:2607.21675v1 [quant-ph] for this version) https://doi.org/10.48550/arXiv.2607.21675 Focus to learn more arXiv-issued DOI via DataCite
人工智能
[AI-0] Explainable Reinforcement Learning for assisting Air Traffic Controllers
链接: https://arxiv.org/abs/2607.22525
作者: Anduel Mehmeti,Gabriella Gigante,Salvatore Venticinque
类目: Artificial Intelligence (cs.AI)
备注: 11 pages (10-page paper plus 1 cover/citation page), 4 figures, 1 table; published in AINA 2025
Abstract:To effectively integrate AI into high-stakes, critical environments such as healthcare, autonomous driving, and aviation–and to advance toward higher levels of automation and seamless human-AI collaboration–building trust in AI-driven solutions is essential. Trust, in turn, is closely linked to the explainability of AI systems. The rapid advancements in AI across various domains have underscored the challenges of establishing trust, raising increasing interest in AI explainability even more when applied to deep learning. In this context, the present work aims to explore the application of explainability techniques to Reinforcement Learning (RL) algorithms, specifically within the safety-critical domain of Air Traffic Control (ATC). Using a simplified ATC environment as an initial testbed, an intelligent agent is trained with a reinforcement learning algorithm to make decisions on alternative flight routes that avoid no-fly zones. As a preliminary explainability approach, a saliency map is employed, providing insights into the input features that most significantly influence the agent’s decision-making process.
[AI-1] he Regression Tax: Decomposing Why Skills Help and Hurt LLM Agents
链接: https://arxiv.org/abs/2607.22520
作者: Darshan Tank,Baran Nama
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Adding procedural skills to an LLM agent is typically evaluated by average improvement in task success. However, this metric hides an important cost: skills can also make agents worse. We measure both sides by comparing agents with and without skills across nearly 6,000 runs spanning two office automation benchmarks and three model harness stacks. This allows us to distinguish two outcomes. A regression is a task solved without skills but failed after skills are added. A residual failure is a task that fails both with and without skills. We find that regressions are substantial enough that the best performing skills outperform others primarily by regressing less, not by gaining more. We identify three causes of regression: (i) skill description osmosis, a skill changes an agent’s behavior simply by being present in context, even when it is never invoked; (ii) grounding displacement, a skill’s prescribed procedure overrides how the agent interprets its inputs; and (iii) verification displacement, where the procedure suppresses checks the agent would otherwise perform on its outputs. Analysing persistent failures reveals the same underlying pattern. Existing skills overemphasize procedural guidance the stage least often responsible for failure while under supporting grounding and verification, the dominant sources of remaining errors. After correcting evaluation artifacts and studying traces, we find many regressions and persistent failures recoverable through better grounding and verification. Procedural skills should be evaluated by decomposing their net effect into gains and regressions, not by aggregate improvement alone. We identify three regression modes skills should avoid, and find that reliability depends more on grounding and verification than on procedural skill choice.
[AI-2] kappa-LoRA: Condition Numbers Reveal Which LoRA Matrices Worth Updating
链接: https://arxiv.org/abs/2607.22489
作者: Jianghui Wang,Silong Yong,Francesco Orabona,Marco Canini,Katia P. Sycara,Yaqi Xie
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Low-Rank Adaptation (LoRA) has become a widely adopted technique for efficient neural network fine-tuning, decomposing model updates into low-rank matrices. However, LoRA remains computationally costly because it updates all matrices uniformly, regardless of their actual contribution to adaptation. This cost is especially prohibitive for large-scale models with billions of parameters and for resource-constrained settings such as edge deployment and on-device fine-tuning. We show for the first time that not all LoRA matrices are equally worth tuning: matrices with smaller condition numbers (the ratio of largest to smallest singular value) are already well-balanced across directions and contribute only marginally to adaptation, whereas matrices with larger condition numbers contain underdeveloped directions that span richer subspaces and drive most of the performance gains. This observation itself is a key contribution of our work, and it motivates a more selective approach to fine-tuning. Building on this insight, we propose \kappa-LoRA, a method that optimizes LoRA by focusing updates on the matrices with the largest condition numbers, which capture the most informative directions of change. By restricting LoRA updates to the top 50% of weight matrices ranked by condition number, \kappa-LoRA halves the trainable parameter count and correspondingly reduces compute and memory cost. Extensive experiments across multiple benchmarks show that this design cuts fine-tuning time by 16.2% on average while matching the accuracy of standard LoRA and reducing memory cost by 4.5%. Further analysis reveals that the condition numbers of the selected matrices consistently decrease over training, suggesting that \kappa-LoRA’s effectiveness stems from targeted spectral rebalancing rather than parameter selection alone.
[AI-3] MineValiCoder: Reliable Code Generation with Test Case Quality Mining and Bipartite Graph-Based Mutual Validation
链接: https://arxiv.org/abs/2607.22471
作者: Zhen Zhao,Qihang Yang,Feifei Dai,Xiangfang Li,Bo Li
类目: oftware Engineering (cs.SE); Artificial Intelligence (cs.AI)
备注: 12 pages, 3 figures, 7 tables
Abstract:Large Language Model (LLM)-based Test-Driven Development (TDD) has advanced automated code generation. However, existing approaches depend heavily on human-crafted test cases and cannot operate effectively when only natural-language requirements are available. Although recent work enables automatic test generation, it often overlooks the inherent stochasticity of LLMs, leading to two key defects: faulty tests generate misleading feedback that distorts code optimization, while mixed-quality test cases produce conflicting evaluation signals that hinder reliable code selection. To address these challenges, we propose MineValiCoder, a collaborative closed-loop TDD framework based on the mutual reinforcement of test-case quality and code quality. MineValiCoder comprises three modules. The Test Case Quality Mining (TCQM) module filters faulty test cases through self-validation, providing reliable optimization supervision. The Parallel TDD Refinement module iteratively optimizes code and generates diverse high-quality code candidates using validated test-case feedback. The Bipartite Graph-Based Code-Test Mutual Validation (BiCoTeV) module dynamically models code-test interactions and performs mutual validation scoring for stable and reliable optimal-code selection. Extensive evaluations across four LLMs and mainstream benchmarks show that MineValiCoder significantly outperforms state-of-the-art methods. Specifically, it achieves Pass@1 scores of 96.34% on HumanEval, 87.40% on MBPP, 64.00% on APPS, and 51.33% on LiveCodeBench. These results demonstrate the effectiveness of MineValiCoder in mitigating LLM stochasticity and improving the reliability of automated code generation. Comments: 12 pages, 3 figures, 7 tables Subjects: Software Engineering (cs.SE); Artificial Intelligence (cs.AI) Cite as: arXiv:2607.22471 [cs.SE] (or arXiv:2607.22471v1 [cs.SE] for this version) https://doi.org/10.48550/arXiv.2607.22471 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[AI-4] Phylogenetic signal in marine mammal and bird vocalizations captured by audio foundation models: the limited benefit of domain-specific pretraining
链接: https://arxiv.org/abs/2607.22458
作者: Víctor Rincón Yepes
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注: 17 pages, 5 figures, 2 supplementary tables. Code, embeddings and derived matrices: this https URL
Abstract:Do learned audio embeddings encode structure that nobody told them to encode? We probe four large pretrained audio models (AST, CLAP, BEATs-bio and BirdNET) with a downstream task none of them saw during training: recovering phylogenetic distance from species vocalizations. If the geometry of the embedding space tracks the tree of life, the representation is picking up something deeper than the labels the model was optimized for. We run Mantel tests across two independent radiations. In 32 marine mammal species (1,754 recordings from the Watkins Marine Mammal Sound Database) the foundation models recover strong phylogenetic signal within the 26 cetaceans (CLAP r=0.82, BEATs-bio r=0.82, AST r=0.74; all p0.001), among the highest acoustic-phylogenetic correlations reported for any taxon. Hand-crafted MFCC features (105d) find nothing (r=0.040, p=0.338). The gap survives after PCA-projecting every embedding down to 105 dimensions, so it is not an artefact of representation size. It also survives a partial Mantel test controlling for dominant frequency (partial Mantel r=0.404, keeping 97% of the variance explained), so it is not just pitch in disguise. We repeat the analysis on 20 bird species using the Jetz et al. (2012) phylogeny, and this time add BirdNET, a classifier trained end-to-end on around 6,000 bird species. The general-purpose foundation models recover the signal again (AST r=0.55, CLAP r=0.52). The unexpected result is that neither BirdNET nor the bioacoustic BEATs-bio beat them (r around 0.32 to 0.36). Matching the training domain to the target taxon does not, by itself, help. Pretrained audio embeddings carry evolutionary information across two independent radiations, and domain-specific pretraining is not required for it to emerge. Comments: 17 pages, 5 figures, 2 supplementary tables. Code, embeddings and derived matrices: this https URL Subjects: Machine Learning (cs.LG); Artificial Intelligence (cs.AI) Cite as: arXiv:2607.22458 [cs.LG] (or arXiv:2607.22458v1 [cs.LG] for this version) https://doi.org/10.48550/arXiv.2607.22458 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[AI-5] Dynamic Capability Scoping for Enterprise AI Agents : A Synthetic Dataset and Three-Source Permission Architecture ICML2026
链接: https://arxiv.org/abs/2607.22445
作者: Halil Burak Noyan
类目: Artificial Intelligence (cs.AI); Cryptography and Security (cs.CR)
备注: Published at the Second Workshop on Agents in the Wild: Safety, Security, and Beyond (AIWILD) at ICML 2026
Abstract:Enterprise AI agents are typically granted static credential sets at configuration time, holding every tool the role might need for every task they perform. This persistent over-privilege expands the attack surface. We argue that capability scoping must follow a dynamic least-privilege principle and be treated as a prevention mechanism before a detection one. A credential that does not exist in an agent’s context cannot be misused regardless of the agent’s reasoning or evasion sophistication. We outline a three-source architecture instantiating this principle: role-based ceilings, a task-context classifier, and policy-derived combination prohibitions creating a layered proactive defense against LLM agent misalignment and misuse cases. The architecture supports both enforcing and observe-only deployment; the latter records agent permission requests inconsistent with task context, producing a behavioral signal usable in misalignment research. As a first step toward evaluating this architecture, we contribute a synthetic dataset of 600 enterprise task prompts grounded in a multi-department company policy, labeled with minimum required permissions across a 15-permission tool-based taxonomy that maps directly to deployable credentials or enforceable guardrails. The dataset is constructed via a two-pass pipeline that separates prompt generation from permission labeling to avoid circularity, and is validated against a 60-record/688 decisions human-reviewed sample (Cohen’s \kappa = 0.917 pre-review and \kappa = 0.967 post-review). Iterating between dataset and policy reduced ceiling violations from 46 to 3, a 93% reduction. This shows that synthetic prompt generation can drive policy refinement when the two are developed together. The dataset, environment specification, and generation pipeline are released to support evaluation of dynamic scoping mechanisms. Comments: Published at the Second Workshop on Agents in the Wild: Safety, Security, and Beyond (AIWILD) at ICML 2026 Subjects: Artificial Intelligence (cs.AI); Cryptography and Security (cs.CR) Cite as: arXiv:2607.22445 [cs.AI] (or arXiv:2607.22445v1 [cs.AI] for this version) https://doi.org/10.48550/arXiv.2607.22445 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[AI-6] Hyperball May Not Be a Free Lunch
链接: https://arxiv.org/abs/2607.22444
作者: Yihao Xiao,Jialong Sun,Zitian Gao,Zeming Wei,Chutian Wang,Ran Tao,Jiaye Teng,Bryan Dai
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注: 14 pages, 4 figures. Code: this https URL . Equal contribution: Yihao Xiao and Jialong Sun. Corresponding author: Bryan Dai
Abstract:For scale-invariant deep networks, Hyperball-style optimizers have shown strong performance in large-scale training by fixing the norms of matrix-valued parameters and normalizing updates. However, the source of their advantage remains unclear. Starting from the angular displacement between consecutive parameter states, we derive an angular effective learning rate that accounts for the parameter-update angle, parameter norm, and update norm. We also show that the conventional norm-based measure is a special case under parameter-update orthogonality. We then decompose optimizer updates into radial and tangential components and analyze how radial updates affect one-step angular displacement. Under the training configurations considered, numerical results show that the radial component has only a limited direct effect on the angular effective learning rate. It therefore cannot explain why MuonH converges more slowly than MuonWD early in training but overtakes it later. To further isolate the underlying mechanism, we devise a heuristic experiment that modifies only the learning-rate schedule so that the dynamics of each optimizer reproduce those of the other. The results suggest that their main difference stems from the evolution of the effective step size rather than an intrinsically superior update direction induced by Hyperball. Our pretraining experiments further show that more aggressive learning-rate decay can accelerate MuonH early in training but may impair its later performance. Thus, maintaining a constant angular velocity does not eliminate the learning-rate-scheduling problem; careful scheduling remains essential to realizing the potential of Hyperball-style optimizers. Our code is publicly available at this https URL.
[AI-7] Robot Learning to Communicate through Projected Visual Abstractions
链接: https://arxiv.org/abs/2607.22434
作者: Danyang Yan,Boyuan Wang,Jiaxun Liu,Boyuan Chen
类目: Robotics (cs.RO); Artificial Intelligence (cs.AI)
备注: Our project website is at: this https URL
Abstract:Humans routinely communicate through abstractions of their bodies, including shadows, silhouettes, and reflections. Yet robots remain largely confined to expressing themselves through their physical morphology. Enabling robots to communicate through such projected visual abstractions requires reasoning not only about bodily motion but also about how that motion is transformed into an external representation perceived by an observer. Among these abstractions, shadows provide a particularly compelling example because they emerge directly from the robot’s embodiment while remaining visually distinct from the body itself. Here, we present a robotic system capable of dynamic shadow expression using a 21-degree-of-freedom dexterous hand with compliant soft skin and a learned shadow self-model. The soft-skinned embodiment reduces light leakage to produce visually continuous silhouettes, while the differentiable self-model learns the mapping between hand configurations and projected shadow appearance through task-agnostic self-exploration. Given a target shadow image or video, the robot optimizes its hand configurations through gradient-based search over 1 the learned self-model and refines the solution through collision-aware simulation to obtain physically feasible motions. For dynamic shadow performance, we further introduce expressive-region objectives, temporal smoothness regularization, and keyframe-based optimization to preserve visually important motion cues while reducing optimization complexity. We demonstrate robotic shadow expression across sign-language gestures, hand-shadow puppetry, and animal motion imitation in both simulation and physical experiments. These results establish a framework for enabling robots to manipulate projected visual abstractions of themselves for communication and visual storytelling.
[AI-8] A Self-Calibrating Agent ic AI Framework for Autonomous Edge Resource Allocation
链接: https://arxiv.org/abs/2607.22400
作者: Fin Gentzen,Marla Grunewald,Iulisloi Zacarias,Mounir Bensalem,Admela Jukan
类目: Networking and Internet Architecture (cs.NI); Artificial Intelligence (cs.AI)
备注: This work has been submitted to the IEEE Transactions on Network and Service Management for possible publication. Copyright may be transferred without notice, after which this version may no longer be accessible
Abstract:Large Language Models (LLMs) are increasingly deployed as autonomous agents, transitioning from static conversational interfaces to dynamic systems capable of complex reasoning, tool execution, and decision-making. However, the operational reliability of these agentic AI systems is fundamentally challenged by the absence of reliable ground truth in open-ended environments and the risk of increasing operational drift over time. To address this challenge, we propose and experimentally evaluate an agentic AI framework, designed to enforce autonomous integrity within LLM-driven systems. We design a self-calibration mechanism that mitigates drift and dynamically approximates ground truth by incorporating an ARIMA forecaster, without requiring continuous human oversight. To demonstrate the effectiveness and reliability of our methodology, we apply it to the complex domain of profiling the resource usage of zero-knowledge workloads in edge computing networks. Experimental results show that the proposed self-calibrating agentic framework successfully profiles the zero-knowledge workloads, achieving a higher accuracy than baseline LLM agents by 91.7% for resource usage prediction and improving the prediction speed by 71.7% compared to pure profiling, establishing a robust foundation for deploying autonomous AI in decentralized infrastructures. Furthermore, the ground truth generation using the proposed ARIMA leaping algorithm is 52% faster than a standard ARIMA forecasting algorithm, while achieving the same accuracy.
[AI-9] HiKV: Hierarchical Importance-Aware KV Cache with Hardware Acceleration for LLM Decoding
链接: https://arxiv.org/abs/2607.22389
作者: Chao Fang,Jun Yin,Man Shi,Marian Verhelst
类目: Hardware Architecture (cs.AR); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: To appear in the IEEE Transactions on Circuits and Systems I: Regular Papers (TCAS-I)
Abstract:With the rapid adoption of long-context large language models (LLMs), the continuously growing KV cache during decoding has become the critical memory bottleneck. To tackle this challenge, we propose HiKV, a novel algorithm-hardware co-design that exploits KV cache redundancy through hierarchical importance awareness. Algorithmically, HiKV compresses the KV cache at two granularities: Stage I evicts unimportant tokens within a fixed budget, and Stage II further loads only the significant elements of each retained token, reaching compression ratios unattainable at a single granularity. Architecturally, we develop a dedicated accelerator centered on a reconfigurable importance sorter that switches between the distinct sorting datapaths each stage requires, unifying the two-stage acceleration in one circuit with minimal overhead. Evaluated on representative LLMs, HiKV achieves up to 7.95x speedup and 90% energy reduction in the attention computation over the vanilla KV cache baseline within negligible 1% accuracy loss. Under iso-accuracy constraints, HiKV outperforms state-of-the-art importance-based methods by achieving an additional 1.82~4.87x reduction in external memory accesses. These benefits are enabled by specialized hardware components that add only 8% to the system area.
[AI-10] Agent ic Root Cause Analysis through Evidence-Grounded Reasoning
链接: https://arxiv.org/abs/2607.22385
作者: Amaury Wei,Olga Fink
类目: Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: 21 pages, 9 figures
Abstract:Diagnosing the root cause of anomalies is essential for safe industrial operation. Despite extensive sensor instrumentation, formulating hypotheses and gathering evidence remains a manual process, creating a major operational bottleneck. While existing data-driven approaches aim to automate this, two critical limitations restrict their deployment: their operate as black boxes unable to justify their diagnosis, and they require scarce labeled examples of faulty operation. To address this gap, we introduce AgentRCA, a zero-shot agentic framework for evidence-grounded root cause analysis. Rather than learning fault-specific mappings, AgentRCA performs inference-time reasoning by combining a data-driven digital twin (modeling normal system dynamics) with a tool-augmented large language model. The agent iteratively gathers statistical evidence, evaluates competing hypotheses, and identifies the physical fault that best explains the observed behavior. Evaluated on a real-world multiphase-flow facility and a large-scale chemical plant, AgentRCA achieves diagnostic performance competitive with fully supervised baselines without relying on fault-specific training. Crucially, it produces transparent reasoning traces that explicitly link observed symptoms to their underlying physical causes. These results establish autonomous hypothesis-driven reasoning as a practical foundation for scalable industrial root cause analysis.
[AI-11] IDEAgent : Agent ic Quality-Diversity Search for Research Idea Generation
链接: https://arxiv.org/abs/2607.22375
作者: Varun Gumma,Navonil Majumder,Soumitra Sinhahajari,Soujanya Poria
类目: Artificial Intelligence (cs.AI)
备注: Under Review
Abstract:Large Language Models (LLMs) have significantly automated the process of scientific discovery over the past few years. However, existing systems share one core limitation: they generate and optimize ideas independently for either Quality or Diversity. This often leads to the generation of ideas in close proximity to one another or to a large set of trivial, unsound, or unclear concepts. In this work, we instead argue that research ideation should be treated as a conjunction of both objectives and framed as a Quality-Diversity (QD) search. In line with this perspective, we introduce IDEAgent, a multi-agent framework that manages the evolution of ideas through lineages. We jointly drive Quality using multi-objective feedback for dedicated repair and refinement, while Diversity is achieved through lightweight sequential memory and explicit comparison against completed ideas, their historical ancestors, and rejected proposals. To systematically evaluate this QD conjunction, we develop Yield, a joint metric that computes the largest set of mutually diverse ideas that satisfy a predetermined quality threshold. Finally, through evaluations across 32 topics spanning 8 domains of Computer Science, we show that IDEAgent outperforms the best baseline by 3.89x on Yield, while achieving non-zero Yield on 8x more topics. We further corroborate these findings through an analysis of quality improvements, showing that repair and refinement are crucial for building logical rigor and clarity while preserving non-obviousness. To encourage future research on QD-search-based ideation, we open-source IDEAgent at this https URL.
[AI-12] Do Agent Benchmarks Measure Capability? Protocol Validity in the Age of Agent ic AI
链接: https://arxiv.org/abs/2607.22368
作者: Jiaqi Shao,Hanck Chen,Wei Zhang,Maxm Pan,Bing Luo
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Agent benchmarks increasingly evaluate repository editing, web research, terminal use, and long-horizon interaction. Their scores support capability claims only when the evaluation protocol keeps the intended capability necessary for success. Recent reward-hacking benchmarks and system reports show that agents can instead recover public solutions, read evaluation artifacts, infer generator structure, manipulate feedback, or benefit from invalid scoring paths; existing responses do not provide a common procedure for attributing these shortcuts and quantifying their effect across benchmarks. We formulate protocol validity and introduce HackDetect, a post-hoc audit that identifies an exposure, determines how the agent used it, and assesses whether the resulting score is misleading. We quantify score inflation with the Mislead gap, defined as the exploit score minus the intended score. We audit 2,385 traces across 15 agent benchmarks and find evidence of exposures and reward hacking in 67.0% of Frontier Science traces and 66.7% of AutoLab tasks. Across paired comparisons, we measure score inflation of 0.45-1.00, showing that benchmark reports should provide evidence that scores reflect the intended capability.
[AI-13] Interior interpretability with attention rollout: contraction and propagation profiles in Transformers
链接: https://arxiv.org/abs/2607.22367
作者: Umberto Biccari,Qian Huang,Enrique Zuazua
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Feature-attribution methods assign scores relating input variables to a model’s output, but do not by themselves characterize how explicitly defined interaction operators compose across its intermediate layers. We introduce \emphinterior interpretability, a propagation-based perspective on internal model organization, and instantiate it for tabular Transformers using attention rollout. We interpret rollout as a row-stochastic operator encoding attention-mediated propagation between feature tokens. By applying classical Doeblin–Dobrushin contraction theory, we show that a rollout operator with a small Dobrushin coefficient is quantitatively close to a rank-one stochastic matrix whose common row is determined by its normalized column sums. This result gives a structural interpretation to the corresponding rollout propagation profile. In Transformers trained for metabolomic age prediction, the measured rollout contraction strengthens with depth. Trained and randomly initialized models also exhibit different propagation profiles, although the present experiments do not establish the predictive relevance of individual rollout-ranked variables. Exploratory comparisons with PCA and GradientExplainer approximations to SHAP reveal localized agreement among highly ranked variables but weak agreement across complete rankings. Attention rollout is therefore used here as a diagnostic of attention-mediated propagation, not as a causal explanation or faithful attribution of the complete Transformer.
[AI-14] Learning Structural Convergence: A Neuro-Symbolic Benchmark for Temporal Reasoning
链接: https://arxiv.org/abs/2607.22365
作者: Michael Romei De Socio,Gian Luca Pozzato,Alessio Merlo
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:High-complexity operational environments require methods that detect and anticipate temporally distributed patterns rather than classify isolated events. This paper introduces TRACTA (Temporal Reasoning and Capability-Trajectory Analysis), a controlled synthetic benchmark for temporal structural reasoning in high-complexity event-driven systems, instantiated through Multi-Domain Operations (MDO)-like scenarios. The benchmark includes three tasks: early_warning, pattern_detection, and run_classification, and compares raw-event neural models, a contract-lite semantic baseline, and a neuro-symbolic configuration operating on semantically grounded trajectories. Results show that raw event-level learning remains informative, but learned temporal modeling over semantic capability and contextual direct-impact trajectories achieves the highest aggregate point estimates, with the largest margins on the temporal tasks. Ablation analysis indicates that capability dynamics, contextual impacts, and temporal structure contribute complementary information. Shortcut diagnostics indicate that the most direct cross-run global-identifier shortcut is controlled in the primary neural input view, while residual shallow signals remain. Overall, the findings support a bounded methodological conclusion: in controlled synthetic settings, semantically grounded trajectories provide an effective representation for temporal structural reasoning, supporting further investigation of semantic interfaces between event data, structured representations, and temporal learning.
[AI-15] Indexing: the Beginning and the End
链接: https://arxiv.org/abs/2607.22361
作者: Alexander Kozachinskiy,Vicente Opazo,Felipe Urrutia
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:We study information bottlenecks in modern deep-learning architectures – RNNs, softmax transformers, linear-attention transformers and state-space models – through the lens of the indexing primitive. In this primitive, the input consists of n bits and one integer i from 1 to n called the index, and the output equals the value of the i -th bit. We introduce causal complexity for masked architectures. We show that architectures with low causal complexity cannot solve the indexing primitive in any constant number of layers when the index appears at the end of the input. In particular, this limitation applies to low-parameter RNNs, SSMs and masked linear-attention transformers. In contrast, small softmax transformers can solve it in one layer, while non-masked linear-attention transformers can solve it in 2, which separates them from their masked counterparts. In turn, when the index appears at the beginning, we show that small RNNs are capable of solving this task in 1 layer, while all the other architectures require 2. All our impossibility results are unconditional and apply even to models that employ infinite-precision real arithmetic. Moreover, experiments for up to n=64 qualitatively align with our theory: configurations with low-parameter theoretical solutions learn the indexing task easily, while configurations that do not admit such theoretical solutions struggle to learn as the sequence length grows. Subjects: Machine Learning (cs.LG); Artificial Intelligence (cs.AI) Cite as: arXiv:2607.22361 [cs.LG] (or arXiv:2607.22361v1 [cs.LG] for this version) https://doi.org/10.48550/arXiv.2607.22361 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[AI-16] achy Mini: Development and Preliminary Evaluation of a Knowledge-Based Generative Social Robot for Higher Education
链接: https://arxiv.org/abs/2607.22345
作者: Stephan Vonschallen,Karim Kaufmann,Dominique Oberle,Friederike Eyssel,Theresa Schmiedel
类目: Robotics (cs.RO); Artificial Intelligence (cs.AI)
备注:
Abstract:Generative social robots (GSRs) powered by large language models offer new possibilities for personalized tutoring in higher education, but also introduce risks related to misinformation, missing transparency, or reinforcing incorrect student responses. Prior work identified knowledge-based design (KBD) requirements that define the informational prerequisites for GSRs to manifest responsible and effective tutoring behavior in higher education. In this paper, we operationalized selected KBD requirements in the Reachy Mini robot platform through system prompting, retrieval-augmented generation, and stateful prompt orchestration. As a result, we present Teachy Mini, a GSR tutoring system that was developed using KBD. To test the system, we conducted a preliminary evaluation study. Participants (N = 24) completed a robot-guided learning session about research methodologies. They learned either with Teachy Mini or with a control version that did not follow KBD principles. Teachy Mini was perceived as significantly more aligned with responsible tutoring behavior than the control robot. Moreover, a manipulation check illustrated that Teachy Mini used personalization, slide-grounded explanations, Socratic questioning, affective support, and learner-anchored feedback more consistently than the control robot. No significant between-condition differences were found in system acceptance, intrinsic motivation, or learning effectiveness, although exploratory analyses suggested a positive effect of KBD on objective learning gains when accounting for learner preferences. Overall, the study offered an initial implementation and preliminary evaluation of KBD for GSR tutoring, indicating that KBD can shape responsible robot behavior and potentially increase learning effectiveness in robot-supported learning.
[AI-17] owards Trustworthy and Cost-Efficient Data Integration: From Naïve RAG to Agent RAG to Agentic RAG
链接: https://arxiv.org/abs/2607.22319
作者: Chuangtao Ma,Arijit Khan
类目: Databases (cs.DB); Artificial Intelligence (cs.AI)
备注: To Appear in the IEEE Data Engineering Bulletin
Abstract:Large language models (LLMs) and AI agents have demonstrated strong potential for data integration in zero-shot and few-shot settings. However, they continue to face significant accuracy and cost challenges in enterprise environments due to a persistent knowledge gap. This paper envisions trustworthy, scalable, and cost-efficient integration through knowledge-grounded LLMs and agents operating within a retrieval-augmented generation (RAG) workflow. Here, trustworthiness refers to evidence-grounded, verifiable reasoning, where integration decisions are transparently supported by retrieved knowledge, robust against hallucination, and consistent across tasks. We trace the evolution from classic RAG to GraphRAG and KG-RAG (knowledge graph-based RAG), highlighting how these paradigms bridge parametric and contextual knowledge. Building on this trajectory, we explore the shift toward Agentic RAG, where autonomous multi-agent systems adaptively plan, retrieve, refine, and reason for complex integration tasks. We examine optimization strategies for cost-efficient integration, addressing computational bottlenecks in large-scale enterprise settings. Finally, we outline open challenges and future directions toward building reliable, explainable, and scalable knowledge-grounded integration systems.
[AI-18] A Roadmap to Impactful Pluralistic Alignment Research
链接: https://arxiv.org/abs/2607.22305
作者: Elinor Poole-Dayan,Jillian Fisher,Atoosa Kasirzadeh,Jacob Andreas,Mitchell Gordon,Michiel A. Bakker
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Pluralistic value alignment—the goal of building AI systems that represent and serve diverse human values and perspectives—has emerged as an active research agenda. Yet, there’s no public evidence that it has shaped the training or evaluation of the AI systems people actually use. We audit the public behavior documents and evaluations of frontier labs, finding none name pluralism as a goal, and as of this writing, no clear indication that production models are explicitly trained or tested for it. This goes against the primary motivations and goals of pluralistic alignment, which revolve around making a positive difference in the models serving billions of users worldwide. We argue that the pluralistic alignment research community should focus on supporting impact and adoption in deployed, widely-used AI systems. We provide evidence for the adoption problem, present three main reasons behind it, and discuss three corresponding areas for future research to address it: 1. The primary justifications for pluralistic alignment so far have been normative or speculative. We need studies showing empirically how pluralistic AI benefits users or society. 2. The pluralistic alignment research community has not settled when pluralistic behavior is warranted or what pluralism ideally looks like in practice. We need to establish a concrete goal for developers to operationalize. 3. Current methods trade off against other desiderata of LLMs in ways that are largely unmeasured, and existing metrics are not “hill-climbable.” We need trade-off-aware evaluations and methods that meet the requirements of production systems. This paper serves as a collective call to action for the pluralistic alignment researchers: progress requires moving beyond normative justification toward empirical foundations, a concrete account of ideal pluralistic behavior, and practical methods and evaluations built for adoption.
[AI-19] AI4PLE: A Methodology for Integrating AI into Product Line Engineering
链接: https://arxiv.org/abs/2607.22260
作者: Bedir Tekinerdogan
类目: Artificial Intelligence (cs.AI)
备注: 14 pages, 4 figures, 1 table
Abstract:Reuse-based development has become increasingly important in the creation of complex systems, offering significant opportunities to reduce costs, improve quality, and accelerate time-to-market. Product Line Engineering (PLE) provides a systematic approach to realizing this potential by enabling the efficient creation, management, and customization of product families by reusing shared assets and capabilities. PLE involves addressing numerous complex decisions, including feature selection, variability management, and configuration optimization, which are critical to the success of a product line. Despite its promise, the systematic integration of Artificial Intelligence (AI) into PLE processes has not yet been comprehensively explored. In this paper, we propose a methodological framework to support the systematic integration of AI into PLE and evaluate its effectiveness through a multi-case study conducted in an industrial context.
[AI-20] IFCLoRA: Topology-Aware Rank Allocation for Parameter-Efficient Fine-Tuning
链接: https://arxiv.org/abs/2607.22251
作者: Wei Zhang,Xinwu Liu,Yihang Cheng
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Low-Rank Adaptation (LoRA) is a widely used parameter-efficient fine-tuning method for large language models, but its performance depends strongly on how a fixed rank budget is distributed across Transformer modules. Existing adaptive-rank methods usually rely on local gradient statistics collected during training, which introduces extra memory and computation and overlooks task-conditioned global information flow. We propose IFCLoRA, a topology-aware rank allocation method applied before fine-tuning. Using a small calibration set and a frozen pretrained model, IFCLoRA builds a sparse task-conditioned interaction graph whose nodes represent LoRA-compatible modules. It combines a global information-flow topology prior with local gradient sensitivity to compute Information-Flow Centrality scores, which estimate each module’s adaptation importance under multi-hop propagation. Ranks are then assigned once under a global budget. Across multiple models, tasks, and low-rank settings, IFCLoRA consistently outperforms LoRA, AdaLoRA, and EVA under matched training configurations and total rank budgets, while retaining training costs comparable to standard LoRA. On mathematical reasoning with LLaMA 3 8B, IFCLoRA improves over LoRA by 1.36 percent at rank 4 and 1.82 percent at rank 8. Further analysis shows task-dependent, non-uniform rank profiles, indicating that global information-flow structure provides an informative and interpretable prior for low-budget parameter-efficient fine-tuning.
[AI-21] Optimization of time-consuming experimental conditions using pseudo-experimental data guided by adaptive polynomial regression
链接: https://arxiv.org/abs/2607.22238
作者: Hirotaka Sugawara,Yujin Taguchi,Kei Minagawa,Yusuke Hiki,Takashi Morikura,Akira Funahashi
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Bayesian optimization (BO) is an optimization method that sequentially proposes the next candidate explainable variables for optimizing target variables by balancing exploration and exploitation. BO is often used under a limited evaluation budget, such as hyperparameter tuning of deep learning. Despite its effectiveness, conventional BO may have poor convergence in practical experimental science where each evaluation is often costly and time-consuming. Recently, BO methods have been proposed that accelerate optimization by using pseudo-experimental data that simulate experimental data. However, when only a limited number of experimental data are available, the generated pseudo-experimental data may be of insufficient quality. In this study, we developed PolyBO to improve optimization time by generating high-quality pseudo-experimental data even when the number of trials is limited. PolyBO performs BO efficiently by generating pseudo-experimental data with an adaptively updated versatile parametric model. This low-capacity polynomial regression model is intended to enable efficient BO even with limited experimental data. PolyBO updates the BO surrogate model with a combined dataset consisting of experimental data and pseudo-experimental data and then performs optimization. Using synthetic benchmark functions with diverse landscapes, we found that PolyBO reduced the optimization time by a median of 42%. For a real-world material composition optimization problem, PolyBO reduced the optimization time by a median of 96% compared with conventional methods. Overall, PolyBO achieves efficient optimization in settings where each experiment requires a long time.
[AI-22] Deconstructing Off-Policy Ratios: Entropy-Scaled Trust Regions for Asynchronous Reinforcement Learning
链接: https://arxiv.org/abs/2607.22186
作者: Guanqun Zhao,Zijun Xie,Binbin Zheng,Enlei Gong,Jiafeng Lu,Yehan Yang,Aoqi Hu,Zeyu Chen
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Asynchronous reinforcement learning (RL) accelerates large language model (LLM) post-training by overlapping rollout generation with policy optimization, but the resulting stale, off-policy data can destabilize optimization and ultimately cause policy collapse. Existing methods typically retain or discard tokens based solely on the magnitude of their importance ratios, applying the same threshold uniformly across token positions. In this work, we reveal that the natural scale of the importance ratio varies systematically with token entropy. Under asynchronous dynamics, this entropy-ratio scaling dictates two distinct phenomena: at low entropy, the inherent train-inference discrepancy is drastically amplified into substantial sampling noise; at high entropy, in-flight weight updates naturally induce pronounced, legitimate exploratory deviations. Consequently, magnitude-only correction inadvertently admits the amplified noise while strictly masking out the essential exploration triggered by in-flight updates. To address this, we propose the Entropy-Scaled Trust Region (ESTR), which scales each token’s off-policy deviation by its local entropy, requiring no auxiliary forward passes or explicit version-switch detection. Across long-horizon agentic tasks and mathematical reasoning benchmarks, ESTR consistently outperforms existing asynchronous methods and achieves the best train-inference consistency. Compared with synchronous GRPO, ESTR attains comparable accuracy while improving training speed by 2.6\times .
[AI-23] Learning Spatiotemporal Decision Priors for Efficient Path Planning under Partial Observability
链接: https://arxiv.org/abs/2607.22166
作者: Yi Liu,Hongda Zhang,Leyao Zou,Chunlei Meng,Ziqing Zhou,Yuning Chen,Zhuo Zou,Lida Xu,Zhongxue Gan,Chun Ouyang
类目: Robotics (cs.RO); Artificial Intelligence (cs.AI)
备注:
Abstract:Path planning under partial observability remains challenging because an agent must make long-horizon navigation decisions from only locally bounded observations. Nevertheless, historical trajectories contain reusable experience-guided directional preferences. Classical planners, however, typically solve each instance from scratch and lack an explicit mechanism to exploit such transferable decision knowledge, often leading to redundant node expansions and locally myopic search behaviors. Motivated by this limitation, this paper proposes ImiPath, a prior-guided learning framework that distills reusable spatiotemporal decision priors from demonstration trajectories and uses them as experience-informed directional guidance to bias planners toward reliable and promising search directions under partial observability. Specifically, ImiPath first constructs a local spatiotemporal observation representation, which encodes the spatial information of the local environment and the temporal information of historical trajectories. The SpatioTemporal-Attention Policy Network (STAPNet) then transforms this representation into dicision priors. These priors are further incorporated into heterogeneous planners as directional guidance, biasing the search toward locally promising regions. Extensive experiments demonstrate that ImiPath achieves competitive path quality and improves search efficiency by reducing redundant node expansions under local observability. Additional physical experiments on a magnetic microrobot platform further validate the adaptability and practical deployment potential of the proposed framework.
[AI-24] Learning on the Job: Continual Learning from Deployment Feedback for Frozen-Weights Agents
链接: https://arxiv.org/abs/2607.22157
作者: Valentin Tablan,Scott Taylor,Kristoffer Bernhem
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:AI agents encounter learning opportunities in every episode they run, and discard nearly all of them: the underlying models are frozen at deployment, so an agent that resolves a difficult request today starts from zero when it recurs tomorrow. Yet ordinary operation already produces feedback, in the form of outcome verdicts and after-the-fact corrections. We show that this feedback is a sufficient signal for continual learning when the frozen model is paired with an external memory that distils each episode into retrievable natural-language rules. On the banking domain of \tau -bench, against a static-RAG control retrieving over the complete policy corpus, learning from the one-bit outcome verdict lifts single-trial success to 1.6 \times the baseline, and learning from corrections to 2.6 \times , converting 22 of the 84 tasks the baseline never solves. The result spans the deployment spectrum, measured on Mistral Large, an open-weights model that organisations with data sovereignty requirements can self-host, and replicated on a frontier model, Claude Sonnet 5. The accumulated memory also transfers: each model, reading the store built by the other, rises above its own no-memory baseline. The harness, protocol, and data are released.
[AI-25] Industrial Tokenization for LLM -Based Health Intelligence: A Federated Architecture for Industrial Evidence Integration
链接: https://arxiv.org/abs/2607.22153
作者: Deshui Li,Xiao-Ming Yuan,Zishun Wang
类目: Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: 10 pages, 1 figure
Abstract:Industrial health management increasingly relies on heterogeneous information sources, including condition monitoring systems, supervisory control and data acquisition systems, maintenance records, inspection results, and prognostic models. Although large language models provide new opportunities for cross-source reasoning, industrial data and analytical outputs differ substantially in structure, temporal resolution, physical meaning, and reliability. Directly integrating such heterogeneous information into a monolithic model may reduce interpretability, traceability, and adaptability to equipment and data changes. This paper introduces Industrial Tokenization, a conceptual interface for transforming source-specific analytical outputs into structured and machine-interpretable units of industrial evidence, termed Industrial Tokens. Unlike numerical tokens used to encode raw time-series data, Industrial Tokens represent domain-grounded evidence together with source, temporal scope, operating context, analytical meaning, quality or confidence information, and provenance. Based on this concept, a federated industrial architecture is proposed, where heterogeneous analytical subsystems retain autonomy while exposing standardized Industrial Tokens to a central reasoning layer. As an initial implementation, this study presents an end-to-end DiagnosisToken pathway based on vibration-diagnostic outputs, rule-based event aggregation, structured textual token generation, and LLM-based interpretation. Other Industrial Tokens, including SCADA-based condition-monitoring tokens, maintenance tokens, and prognostic tokens, are reserved as future extensions. The proposed framework positions Industrial Tokenization as a semantic interface between domain-specific industrial intelligence and LLM- or agent-based reasoning, rather than another method for encoding raw industrial data.
[AI-26] riGlue: a Biology-Inspired Generative Model for Generating Molecular Glue-Induced Ternary Complex
链接: https://arxiv.org/abs/2607.22143
作者: Yuliang Yan,Shuo Yan,Haochun Tang,Yiqin Sun,Enyan Dai
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Molecular glue degraders have emerged as a promising strategy for targeted protein degradation by inducing ternary complex formation between an E3 ubiquitin ligase and a target protein. Despite their therapeutic potential, computational design of molecular glues remains largely unexplored. Unlike conventional structure-based drug design, molecular glue design is governed by the unknown protein-protein interface and requires the simultaneous modeling of ligand generation, protein-protein docking, and ternary complex assembly. In this work, we formulate molecular glue design as a ternary complex generation problem and propose a biology-inspired generative framework, TriGlue. Motivated by the mechanism of molecular glue action, we decompose ternary complex generation into two coupled stages: interface estimation and interface-conditioned complex generation. First, we develop an SE(3)-equivariant interface estimation module that predicts a geometrically constrained protein-protein interface from unbound monomer structures. Second, we introduce an interface-conditioned ternary flow matching network that jointly generates the molecular glue and predicts the rigid-body transformation required to assemble the ternary complex. Extensive experiments demonstrate that TriGlue generates chemically valid molecules and produces plausible ternary complexes, which highlight the potential of biology-inspired generative modeling for accelerating molecular glue discovery. Our code is available at this https URL.
[AI-27] One Hand Watches The Other: Dynamic Multi-Agent Cooperation for Sample-Efficient Bimanual Manipulation in Dynamic Environments
链接: https://arxiv.org/abs/2607.22119
作者: Jan Ole von Hartz,Abhinav Valada,Joschka Boedecker
类目: Robotics (cs.RO); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注:
Abstract:Multi-stream robot manipulation policies achieve unparalleled sample efficiency and generalization by modeling actions relative to environmental reference frames. However, existing approaches typically assume these frames to be strictly exogenous. This causal assumption collapses in dynamic settings, such as when a single robot arm manipulates a moving object or when two arms coordinate, where each arm effectively becomes part of the dynamic environment of the other. We propose DynaMAC, a lightweight, policy-agnostic framework that resolves this causal limitation while preserving the sample efficiency, computational speed, and flexibility of multi-stream policies, DynaMAC treats the opposite arm as a dynamic task parameter, thereby providing a unified formulation for dynamic manipulation and bimanual coordination without requiring an explicit leader-follower relationship. To rigorously evaluate these capabilities, we introduce DynaBench, a novel benchmark for robot manipulation in dynamic environments. Across both dynamic environments and bimanual manipulation tasks, DynaMAC outperforms leading probabilistic and generative baselines by over 35 percentage points while requiring 20 times fewer samples. Crucially, DynaMAC generalizes zero-shot from static demonstrations to dynamic environments, substantially simplifying data collection and establishing an elegant bridge toward human-robot collaboration.
[AI-28] Benchmarking Text-to-SQL under Role-Based Access Control SIGMOD2027
链接: https://arxiv.org/abs/2607.22115
作者: Yang Fei,Yangfan Jiang,Yin Yang,Xiaokui Xiao
类目: Databases (cs.DB); Artificial Intelligence (cs.AI)
备注: Accepted at ACM SIGMOD 2027
Abstract:Given a database S and a natural language question Q, text-to-SQL systems aim to generate an SQL query that correctly answers Q when executed against S. Currently, popular text-to-SQL benchmarks mostly assume unrestricted access to S; in practice, however, user access is often restricted, e.g., through role-based access control (RBAC) policies. This leads to a potential disconnect between benchmarking results and real-world performance: an LLM with high benchmark scores might perform poorly in an access-controlled environment, by frequently violating RBAC, or rejecting a query q that could be answered with only permitted data in S. Motivated by this, we present a comprehensive text-to-SQL benchmarking framework with realistic RBAC constraints, which features an LLM-assisted workflow that augments existing text-to-SQL benchmarks with plausible user roles and access policies. To do so, we formulate the problem of role synthesis as a structured reasoning process over the database schema, in which the LLM first infers the application context from the schema, and then derives role responsibilities and access scopes consistent with this context. This process is audited by human-in-the-loop quality control, in which domain experts perform metric-guided screening on the generated roles. Besides the augmented dataset, the proposed framework also contains evaluation metrics that identify RBAC-specific failure modes, and disentangle SQL utility from access-control compliance. We apply the proposed framework to several widely-used benchmarks, and conduct a systematic empirical study of state-of-the-art text-to-SQL systems. The results show that many solutions (especially open-weight LLMs) with high benchmarking scores under an unrestricted setting suffer sharp performance degradation once access constraints are in place, due to frequent RBAC violations.
[AI-29] Reasoning Denoiser: Denoising Reasoning Reasoning Traces for Hallucination Detection in Large Reasoning Models
链接: https://arxiv.org/abs/2607.22098
作者: Junlin Fang,Do Nguyen-Thanh,Xiaogang Xu,Zhen Fang,Sean Du
类目: Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注:
Abstract:Large reasoning models (LRMs) generate long reasoning traces before producing final answers. While these traces may contain useful signals for hallucination detection, harnessing them is non-trivial because long trajectories often include noisy steps that obscure the cues relevant to truthfulness assessment. In this paper, we identify two prevalent forms of reasoning noises, i.e., irrelevant steps and repetitive steps, and show that both substantially degrade hallucination detection performance. Existing confidence-based scores and naive embedding-based filtering fail to reliably separate noisy from informative steps. To address this challenge, we propose REDE, a novel learning framework for denoising reasoning traces for hallucination detection. Specifically, REDE leverages final-answer attention as an automatic supervision signal to shape the step-level representation space, yielding refined embeddings in which noisy steps can be reliably identified and filtered. REDE can be readily plugged into diverse hallucination detectors by operating on the filtered reasoning trajectory after removing noisy steps. Extensive experiments on multiple reasoning benchmarks show that REDE consistently improves detection performance over competitive baselines.
[AI-30] A Leakage-Free Stacked Ensemble Method for Multiclass Classification
链接: https://arxiv.org/abs/2607.22081
作者: S. P. Sharmila,Aruna Tiwari
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注: 7 pages, 3 figures, This paper is presented at IEEE World Congress on Computational Intelligence (WCCI 2026) June 2026, Maastricht, The Netherlands
Abstract:Multiclass classification is a fundamental problem across a wide range of domains. It is still challenging due to possession of high inter-class similarity, class imbalance datasets, and variability in data distributions. Rule-based classifiers such as XGBoost often achieve stronger performance on structured features, but they are limited in capturing smooth functional relationships among variables. Similarly, neural network models can represent complex nonlinear interactions but frequently suffer from overfitting and generalization issues. To address these limitations, we propose LFS-FRAME, a Leakage-Free Stacked ensemble framework that integrates functional learning using Kolmogorov-Arnold Networks (KAN) and rule-based learning via XGBoost for robust multiclass classification. The proposed framework constructs unbiased meta-features by employing a strict out-of-fold stacking strategy to ensure complete isolation between training and validation data hence preventing performance leakage. By learning over probabilistic outputs from heterogeneous base learners, the meta-classifier effectively exploits both global functional patterns and sharp decision boundaries present in the complex data. Experimental evaluations on multi-class datasets demonstrate that LFS-FRAME improves performance metrics, and overall accuracy is 89.85% in identifying major families and 81.74% in identifying sub-families relative to strong single-model baselines. These results highlight the effectiveness of leakage-free functional and rule-based stacking for reliable and generalizable multiclass classification. Comments: 7 pages, 3 figures, This paper is presented at IEEE World Congress on Computational Intelligence (WCCI 2026) June 2026, Maastricht, The Netherlands Subjects: Machine Learning (cs.LG); Artificial Intelligence (cs.AI) Cite as: arXiv:2607.22081 [cs.LG] (or arXiv:2607.22081v1 [cs.LG] for this version) https://doi.org/10.48550/arXiv.2607.22081 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[AI-31] CEL: Comprehensive Counterfactual Explanations Library and Benchmark KDD
链接: https://arxiv.org/abs/2607.22045
作者: Oleksii Furman,Łukasz Lenkiewicz,Marcel Musiałek,Maciej Zięba
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注: 16 pages, 5 figures. Accepted for presentation at the XKDD and Beyond Workshop (non-archival)
Abstract:Counterfactual explanations are a prominent approach in explainable artificial intelligence (xAI), providing actionable guidance on what input changes would alter a model’s prediction to a desired outcome. While early methods primarily focused on minimal feature changes, recent work incorporates additional properties such as sparsity, actionability and plausibility. Despite this progress, fair and systematic evaluation remains challenging. Existing studies often rely on different data splits, predictive models, and evaluation metrics, which limits objective comparison across methods. To fill this gap, we introduce CEL (Counterfactual Explanations Library), a unified library and benchmark for counterfactual explanations designed to support consistent implementation and evaluation. CEL includes 18 datasets of varying size and complexity and provides implementations or reimplementations of 14 widely used counterfactual methods. Using this standardized setup, we conduct a comprehensive quantitative comparison across a variety of methods on datasets that differ in size, number, and types of attributes. The evaluation protocol incorporates multiple complementary metrics capturing validity, coverage, sparsity, proximity, and distributional plausibility, including density- and outlier-based measures to assess the realism of generated counterfactuals. To the best of our knowledge, this is the first comprehensive benchmark that systematically evaluates recent counterfactual explanation methods within a unified and reproducible framework. While prior libraries and benchmarking efforts exist in the literature, many are outdated, limited in scope, or lack consistent evaluation protocols. The proposed benchmark aims to improve reproducibility, enable fair comparison, and establish a workbench for the development of future counterfactual explanation methods.
[AI-32] Sparse by Command: Task-Conditional Compute Skipping for Multi-Task Inference Accelerators MICRO MICRO2026
链接: https://arxiv.org/abs/2607.22038
作者: Afzal Ahmad,Gaoyu Mao,Shoubo Hu,Hui-Ling Zhen,Mingxuan Yuan,Xinyu Chen,Wei Zhang
类目: Hardware Architecture (cs.AR); Artificial Intelligence (cs.AI)
备注: Accepted at IEEE/ACM International Symposium on Microarchitecture (MICRO 2026)
Abstract:Multi-task inference models share a single backbone across diverse tasks, yet execute identical computation regardless of which task is active - wasting energy and cycles on task-irrelevant operations. We observe that the task command, typically available before inference begins, provides a free signal that can be exploited to skip unnecessary computation at the hardware level. We present a HW/SW co-designed approach in which a lightweight gating network, trained jointly with the backbone, predicts per-tile binary execution masks conditioned on the task input. Each tile corresponds to a fixed group of output channels (the native scheduling granularity of the accelerator), enabling masked tiles to be skipped with zero overhead. This yields a task-dependent reduction in compute, where each command activates only the subset of the network it requires, without changes to the model architecture or inference pipeline. We co-design the full system stack: a command-conditioned training procedure that learns hardware-aligned tile masks under a sparsity objective; an instruction set architecture whose instructions carry per-tile bitmask fields, allowing the hardware to skip masked tiles without software intervention; and a tiled inference accelerator with configurable parallelism, double-buffered memory, and INT8 datapath that natively supports sparse tile execution. We prototype on an AMD/Xilinx Alveo U50 FPGA and evaluate on a closed-loop visuomotor driving task in CARLA autonomous driving simulator. Task-conditional sparsity reduces FLOPs by 66-76% while maintaining driving quality. On-device latency decreases by 51-59%, from 9.12 ms to 3.74-4.44 ms (2.1-2.4x speedup), with energy per inference dropping from 263 to 108-128mJ. Comments: Accepted at IEEE/ACM International Symposium on Microarchitecture (MICRO 2026) Subjects: Hardware Architecture (cs.AR); Artificial Intelligence (cs.AI) Cite as: arXiv:2607.22038 [cs.AR] (or arXiv:2607.22038v1 [cs.AR] for this version) https://doi.org/10.48550/arXiv.2607.22038 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[AI-33] Agent Security Needs Redefinition through a Holistic Framework ICML2026
链接: https://arxiv.org/abs/2607.22024
作者: Vincent Siu,Jingxuan He,Kyle Montgomery,Zhun Wang,Chenguang Wang,Dawn Song
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI)
备注: ICML 2026 Position Paper
Abstract:Agent security is widely treated as a question about action content. Defenses ask whether an instruction looks malicious. Benchmarks ask whether an agent performs a harmful sounding action. \textbfWe argue that agent security is fundamentally a contextual problem, and that the current content based framing systematically misdefines it. A command to ``delete user data’’ might be a routine administrative request or a prompt injection attacking production systems, and the content alone cannot distinguish the two. Authorization context can. Across every injection task in AgentDojo and WASP, the same action is one an authenticated user would plausibly request in a routine workflow, which makes the conflation a structural property of evaluating security through content. We operationalize contextual security through four properties that must hold jointly and be evaluated continuously across the agent’s trajectory. Source Authorization asks who issued the command. Task Alignment specifies the agent’s authorized objective. Action Alignment evaluates whether each action serves that objective. Data Isolation governs information flows across privilege boundaries. Under this reframing, indirect prompt injection becomes a Source Authorization violation. Snapshot benchmarks are structurally incapable of evaluating Data Isolation. Existing defenses are reorganized around the property they actually approximate. The contextual reframing changes which defenses are coherent, which evaluations measure something useful, and which attack patterns evaluation can see at all. Comments: ICML 2026 Position Paper Subjects: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI) Cite as: arXiv:2607.22024 [cs.CR] (or arXiv:2607.22024v1 [cs.CR] for this version) https://doi.org/10.48550/arXiv.2607.22024 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[AI-34] Practical Graph Optimisation and AI-Driven Models for Active Directory Security Hardening
链接: https://arxiv.org/abs/2607.22009
作者: Huy Q. Ngo
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI)
备注: PhD Thesis
Abstract:Microsoft’s Active Directory (AD) is a directory service that enables the IT admin to manage security permissions and control access within a Windows domain network. As a core management system in many of organisation, AD has become a primary target for adversaries. While many solutions for hardening attack graphs exist, these efforts fall short in addressing several key practical challenges specific to the AD attack graph. First, existing models often assume the graph is static, whereas a real-world AD environment is highly dynamic. Second, most proposed solutions are limited to the defensive measure of revoking vulnerabilities (edge removal), while more active defence mechanisms are largely unstudied. Third, because not all remediations are implementable, a practical end-to-end model must incorporate system admin feedback into the prioritisation process. This thesis aims to address these limitations by studying and proposing a number of game-theoretic and optimisation-based decision-making models. First, we propose a honeypot/decoy placement model based on the principle of minimising the number of shortest paths and the number of Domain Admin-reachable nodes. Second, building on this model, we introduce a defence strategy that considers the dynamic/temporal nature of the AD graph, where the objective is to find the location to deploy decoys that maximises the worst-case incident response time. Third, we introduce an adaptive prioritisation model that queries each high-risk attack path to the IT administrator for mediation. Finally, we introduce an end-to-end adaptive prioritisation model that minimises the approval effort of the system admin by finding a general adaptive edge-removal policy that generalises the system admin’s decisions to edges with similar risk features. We show that the problems underlying all of the contributed models are computationally intractable.
[AI-35] Learning as Reasoning Unfolds: Progressive Rollout Allocation for Efficient Reinforcement Learning
链接: https://arxiv.org/abs/2607.22002
作者: Heyang Jiang,Henry Liu,Baharan Mirzasoleiman
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Reinforcement learning with verifiable rewards (RLVR) has emerged as a highly effective framework for improving LLM reasoning, with methods such as GRPO among its most successful instantiations. However, GRPO relies on repeated generation of long chain-of-thought rollouts. Training time scales with the number of rollouts, a large fraction of which are uninformative. Thus, GRPO is computationally expensive and unstable. To mitigate this, existing approaches either generate a larger pool of rollouts and filter the most informative prompts, or leverage historical signals for filtering at later stages of training. These strategies offer modest performance gains, but slow down the overall process. To address this, we propose VarIance Guided Online Rollout allocation (VIGOR) which instead of allocating a fixed rollout budget per example, begins with a small number of rollouts for all examples in a batch and iteratively allocates additional rollouts to those with the highest group reward variance until a fixed total rollout budget is reached. Theoretically, we show that under RLVR, reward variance controls the gradient magnitude, and derive VIGOR’s closed-form speedup ratio over GRPO, which grows with refinement rounds under Pareto-distributed reward variance. Experiments on mathematical reasoning and coding tasks show that VIGOR reaches target accuracy with up to 2.3 \times fewer rollouts on math, reaches GRPO’s final coding full pass rate with 1.49 \times fewer rollouts, and improves the coding average test pass rate by 3.4 points.
[AI-36] From Perturbation Correction to Geometry-Aware Sampling: Sharpness-Guided Equilibrium Sampling for Balanced Flat Minima in Long-Tailed Learning
链接: https://arxiv.org/abs/2607.21999
作者: Jiaxin Deng,Junbiao Pang
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Long-tailed learning couples two sources of poor generalization: head classes dominate training exposure, while under-represented classes often converge to sharper regions of the loss landscape. Conventional re-sampling addresses the former without considering geometry, whereas existing long-tailed sharpness-aware minimization (SAM) methods modify losses or perturbations only after biased mini-batches have been drawn. We introduce Sharpness-Guided Equilibrium Sampling (SGS), which treats the sampling distribution as an active control variable for optimization geometry. SGS dynamically adjusts subsequent mini-batches by increasing the sampling probability of less frequently sampled classes while suppressing classes with large SAM-induced loss changes, using only cumulative class counts and EMA sharpness estimates obtained from the standard SAM update, without class-wise perturbations or additional backward passes. We characterize this sampling process through a continuous-time stochastic differential equation and a sampling-dependent PAC-Bayes analysis, explaining how frequency-sharpness feedback can move training toward a more balanced flatness profile. On CIFAR-100 LT with an imbalance ratio of 100, SGS-SAM improves Focal-SAM by 10.85 points in tail accuracy and 3.56 points overall. On ImageNet-LT, it improves ImbSAM by 6.59 points on tail classes and 1.20 points overall. Its training time is only 1.02\times that of vanilla SAM. Beyond these gains, SGS establishes a sampling-side route to loss-landscape control, suggesting that future long-tailed methods can jointly regulate data exposure and optimization geometry rather than treating either as fixed.
[AI-37] Unified Static-Dynamic Pruning for Efficient LLM Inference VLDB
链接: https://arxiv.org/abs/2607.21985
作者: Jinhyeok Kim,Yejoon Lee,Jaeyoung Do
类目: Distributed, Parallel, and Cluster Computing (cs.DC); Artificial Intelligence (cs.AI); Hardware Architecture (cs.AR); Machine Learning (cs.LG)
备注: 14 pages, 13 figures. Accepted for publication in Proceedings of the VLDB Endowment (PVLDB), Volume 19, Issue 11, 2026. Code: this https URL
Abstract:The increasing deployment of large language models (LLMs) has magnified the computational and memory bottlenecks of autoregressive decoding, where low compute intensity and bandwidth-bound kernels dominate inference cost. Weight pruning offers a promising remedy, but existing methods remain confined to either static pruning (SP), which permanently removes redundant weights but lacks adaptivity, or dynamic pruning (DP), which adapts to input sparsity but introduces runtime irregularity. This paper presents SPDP, a unified sparse-inference framework that integrates unstructured SP with input-adaptive DP for efficient LLM inference on GPUs. SPDP co-designs a new Tiled-Column-wise Bitmap Compressed (Tiled-CBC) format and two complementary GPU kernels: (1) a CUDA-core spMspV kernel featuring Hybrid Activation-aware Dynamic Shared-Memory Bitmap Decoding (HAD-SMBD) for fine-grained, runtime activation skipping, and (2) a Tensor-Core SpMM kernel optimized for prefill computation. This joint format-kernel design harmonizes static and dynamic sparsity, maintaining bandwidth-efficient memory access and high compute intensity under both phases of LLM inference. Comprehensive evaluations on inference-optimized GPUs demonstrate that SPDP achieves 1.24x-1.37x average speedup (up to 2.51x) over state-of-the- art sparse frameworks such as SpInfer, while matching. perplexity with up to 25% higher sparsity. SPDP advances the inference efficiency-quality Pareto frontier, showing that unified static-dynamic pruning can deliver substantial throughput and performance-per-watt improvements in large-scale LLM serving Comments: 14 pages, 13 figures. Accepted for publication in Proceedings of the VLDB Endowment (PVLDB), Volume 19, Issue 11, 2026. Code: this https URL Subjects: Distributed, Parallel, and Cluster Computing (cs.DC); Artificial Intelligence (cs.AI); Hardware Architecture (cs.AR); Machine Learning (cs.LG) Cite as: arXiv:2607.21985 [cs.DC] (or arXiv:2607.21985v1 [cs.DC] for this version) https://doi.org/10.48550/arXiv.2607.21985 Focus to learn more arXiv-issued DOI via DataCite (pending registration) Submission history From: Jinhyeok Kim [view email] [v1] Fri, 24 Jul 2026 05:19:41 UTC (1,740 KB) Full-text links: Access Paper: View a PDF of the paper titled Unified Static-Dynamic Pruning for Efficient LLM Inference, by Jinhyeok Kim and 2 other authorsView PDFTeX Source view license Current browse context: cs.DC prev | next new | recent | 2026-07 Change to browse by: cs cs.AI cs.AR cs.LG References Citations NASA ADSGoogle Scholar Semantic Scholar export BibTeX citation Loading… BibTeX formatted citation loading… Data provided by: Bookmark checked="checked"class=“labs-tab-input”> Bibliographic Tools Bibliographic and Citation Tools Bibliographic Explorer Toggle Bibliographic Explorer (What is the Explorer?) Connected Papers Toggle Connected Papers (What is Connected Papers?) Litmaps Toggle Litmaps (What is Litmaps?) scite.ai Toggle scite Smart Citations (What are Smart Citations?) Code, Data, Media Code, Data and Media Associated with this Article alphaXiv Toggle alphaXiv (What is alphaXiv?) Links to Code Toggle CatalyzeX Code Finder for Papers (What is CatalyzeX?) DagsHub Toggle DagsHub (What is DagsHub?) GotitPub Toggle Gotit.pub (What is GotitPub?) Huggingface Toggle Hugging Face (What is Huggingface?) ScienceCast Toggle ScienceCast (What is ScienceCast?) Demos Demos Replicate Toggle Replicate (What is Replicate?) Spaces Toggle Hugging Face Spaces (What is Spaces?) Spaces Toggle TXYZ.AI (What is TXYZ.AI?) Related Papers Recommenders and Search Tools Link to Influence Flower Influence Flower (What are Influence Flowers?) Core recommender toggle CORE Recommender (What is CORE?) Author Venue Institution Topic About arXivLabs arXivLabs: experimental projects with community collaborators arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website. Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them. Have an idea for a project that will add value for arXiv’s community? Learn more about arXivLabs. Which authors of this paper are endorsers? | Disable MathJax (What is MathJax?) mathjaxToggle(); We gratefully acknowledge support from our major funders, member institutions, , and all contributors. About Help Contact Subscribe Copyright Privacy Accessibility Operational Status (opens in new tab) Major funding support from
[AI-38] ACME: A Multi-Cultural Multi-Embodiment Social-Navigation Dataset
链接: https://arxiv.org/abs/2607.21964
作者: Shashank Rao Marpally,Allan Wang,Atharva Ghotavadekar,Renato Alexandre Ribeiro,Nhat Le,Pilar Bachiller-Burgos,Pranav Goyal,Subham Agrawal,Yasuhiro Nitta,Howard Ziyu Han,Daeun Song,Masaki Kuribayashi,Kohei Uehara,Xiyue Wang,Yangzhe Kong,Duc M. Nguyen,Amirreza Payandeh,Gerardo Pérez-González,Alejandro Torrejón-Harto,Jeeho Ahn,Tisha Jain,Andrew Stratton,Elvin Yang,Jorge de Heuvel,Nico Ostermann-Myrau,Sai Anudeep Sajja,Mithilya Raj,Daisuke Sato,Gaston Rouquette,Nikolas Martelaro,Maki Sugimoto,Hironobu Takagi,Chieko Asakawa,Maren Bennewitz,Aaron Steinfeld,Xuesu Xiao,Christoforos Mavrogiannis,Harold Soh
类目: Robotics (cs.RO); Artificial Intelligence (cs.AI)
备注: 24 Pages, 19 Figures, Submitted to IJRR on June 29th 2026
Abstract:Understanding how robots and humans move in shared spaces is essential for designing effective social robot navigation policies and predicting human behavior. However, existing datasets often lack the diversity needed to capture differences in culture, geography, and human-robot interaction-factors that strongly shape appropriate social behavior. To address this gap, we introduce ACME: A Cross-cultural, Multi-Embodiment dataset for social navigation. A large-scale data collection effort across 8 sites in 5 countries, using 7 robot embodiments, ACME is a large and diverse multi-modal dataset aimed at advancing social navigation research, providing 29.35 hours of onboard robot data and 43.5 hours of overhead pedestrian tracking data. Unlike prior datasets, it focuses on capturing goal-driven social navigation behavior in complex social scenarios with explicit robot-crowd interaction through robot speech. To facilitate learning navigation policies and predicting pedestrian trajectories, ACME provides 3D and 2D scene features, odometry, interaction information, and human-annotated pedestrian trajectory labels. We make ACME easy to use by providing both human-readable data for each sensor modality as well as raw binary data. Our qualitative and quantitative analyses show that our dataset captures more challenging scenarios and a broader distribution of pedestrian behavior than previous datasets.
[AI-39] MA-DAR: Manifold-Aligned Dynamic Adaptive Routing for Continual Temporal Knowledge Graph Reasoning
链接: https://arxiv.org/abs/2607.21949
作者: Xiangjun Shi,Chong Mu,Jinchuan Zhang,Lizong Zhang,Yuefeng He,Shang Liu
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Continual temporal knowledge graph (TKG) reasoning aims to continuously incorporate newly emerging facts while preserving previously acquired knowledge. Replay-based continual learning has achieved promising performance by revisiting historical representations. However, existing methods primarily focus on what to replay, while largely overlooking how replayed representations should be integrated with current ones. Such direct integration often gives rise to two critical forms of representation conflict: \textitnorm domination and \textitsemantic blurring, ultimately degrading continual reasoning performance. To address these challenges, we propose MA-DAR (Manifold-Aligned Dynamic Adaptive Routing), a lightweight plug-and-play framework for replay representation fusion. MA-DAR first aligns replayed and current representations onto a shared manifold to alleviate distribution discrepancies. It then employs a dynamic gating mechanism to learn dimension-wise fusion weights, adaptively determining the contribution of replayed and current representations to the fused representation. Furthermore, a polarization regularizer encourages more decisive routing behaviors by discouraging ambiguous gating decisions, resulting in more stable and effective knowledge integration. Extensive experiments on four public continual TKG benchmarks demonstrate that MA-DAR consistently improves the performance of representative TKG encoders while remaining effective under different replay settings. Comprehensive ablation studies and visualization analyses further verify the effectiveness of manifold alignment and dynamic adaptive routing in mitigating representation conflicts and improving continual reasoning.
[AI-40] Semiotic logical hexagon theory for LLM logical reasoning
链接: https://arxiv.org/abs/2607.21933
作者: Yunyao Zhang,Xinglang Zhang,Zeliang Chen,Junqing Yu,Zikai Song
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Large language models (LLMs) have become powerful tools for language understanding and logical reasoning. However, they still make mistakes when a problem requires both understanding meaning and following logic. A key reason is that natural-language statements often carry implicit semantic relations before any formal reasoning begins. If these hidden meanings are not properly organized, the model may reach incorrect conclusions even when the subsequent reasoning process appears logically valid. Existing methods improve reasoning through decomposition, symbolic translation, external solvers, or self-verification, but pay comparatively less attention to the semantic structure on which reasoning depends. In this paper, we further investigate how semantic organization influences logical reasoning in LLMs. To this end, we propose HexLogicAgent, a framework that first organizes the meaning of natural-language statements and then guides logical reasoning through structured verification. In our investigation, we also make two observations. First, incomplete semantic representations, rather than deductive inference itself, are a major source of logical reasoning failures in LLMs. Second, explicitly modeling the complete structure of semantic opposition substantially delays the degradation of reasoning performance as logical complexity increases. Experiments on challenging logical reasoning benchmarks demonstrate that HexLogicAgent consistently improves reasoning reliability across multiple LLMs. The core idea is supported by a logical hexagon theory, which explains why a complete structure of opposing meanings is necessary for reliable reasoning.
[AI-41] Generalized Neural Operator for Parametric and Boundary-Value Problems
链接: https://arxiv.org/abs/2607.21932
作者: Ruoyan Li,Yizhou Sun,Wei Wang
类目: Computational Engineering, Finance, and Science (cs.CE); Artificial Intelligence (cs.AI)
备注:
Abstract:Developing foundational neural simulators for Partial Differential Equations (PDEs) requires robust generalization across diverse physical parameters and boundary conditions. However, current deep learning approaches largely face a structural trade-off between condition-agnostic deployment and physical fidelity. Purely data-driven operators infer the underlying physics implicitly and thus lack the explicit constraints needed to ensure physically valid solutions across varying domains, rendering the learning problem ill-posed. On the other hand, Physics-Informed Neural Networks (PINNs) enforce rigorous physical constraints but necessitate costly, instance-specific optimization. Furthermore, the massive scale of emerging foundational operators has severely degraded their inference speeds, making them computationally uncompetitive with traditional numerical solvers. To address this bottleneck between condition-agnostic deployment, physical rigor, and inference efficiency, we propose a \textitGeneralized Neural Operator. By formalizing the classical conditions for well-posedness within neural operators, our framework demonstrates the theoretical benefits of explicitly conditioning on PDE parameters and boundary conditions. To implement this synthesis without compromising computational speed, we introduce three novel architectural components: a parameter-gated mixture of kernels for efficient parameter generalization, a generalized boundary transfer operator that projects arbitrary boundary constraints into a unified latent Dirichlet representation, and a specialized training objective to ensure stability. Extensive experiments demonstrate that our theoretically grounded approach achieves superior generalization across heterogeneous physical regimes while maintaining strict inference efficiency comparable to conventional numerical baselines.
[AI-42] RW: TRACE-RealWorld—An Auditable Consistency Contract for World Models as Materialized Views
链接: https://arxiv.org/abs/2607.21910
作者: Edward Y. Chang
类目: Artificial Intelligence (cs.AI); Databases (cs.DB)
备注: 50 pages, 22 tables, 10 figures
Abstract:TRACE-RealWorld addresses a core data-management problem: maintaining an actionable materialized view over a continuously changing physical world when reads of the base state are priced, delayed, heterogeneous, and fallible. Its data-management contributions are a commitment-level validity abstraction for materialized predictions; consequence-conditioned adaptive view maintenance; transaction-style, dependency-scoped compensation for commitments invalidated after authorization; and append-only provenance supporting exact replay. The work builds directly on materialized-view maintenance, adaptive stream synchronization, transaction recovery, sagas, data freshness, and provenance. The end-to-end Flood-SAR evaluation treats sensing as physical data acquisition and measures freshness, verification cost, stale reads, recovery scope, restoration failure, and replayability through six pre-registered questions with held-out seeds. The contribution is therefore not a new predictive model, but a consistency, recovery, and accountability contract for deploying learned world representations as operational data systems.
[AI-43] Multi-Agent System-driven Digital Twins for predictive maintenance: architectures technologies and open research challenges
链接: https://arxiv.org/abs/2607.21873
作者: Korota Arsène Coulibaly,Mohamed Hamlich
类目: Artificial Intelligence (cs.AI)
备注: 33 pages, 5 figures, 4 tables, paper submitted to Internet of Things Journal(Elsevier)
Abstract:Digital twins have emerged as a foundational technology within the context of Industry 4.0, offering a paradigm for the real-time virtual representation of physical systems. However, managing their growing complexity, particularly in distributed industrial environments, requires intelligent architectures capable of autonomous decision-making, dynamic adaptability, and inter-agent coordination. This systematic review explores the intersection between Multi-Agent Systems and Digital Twins, with a particular focus on predictive maintenance applications in resource-constrained contexts. Through a critical analysis of over 547 papers published in high-impact journals (IEEE Transactions, Nature, Elsevier, MDPI), we establish a taxonomy of existing hybrid architectures, identify persistent technological bottlenecks, and formulate three open research questions concerning: (i) the deployment of artificial intelligence on resource-constrained microcontrollers, (ii) distributed multi-node coordination via lightweight communication protocols, and (iii) the hierarchical orchestration of Digital Twins toward smart factory control integrating residual life estimation and explainable Artificial Intelligence. The results of this analysis reveal that, despite significant progress, no existing system offers an integrated embedded-distributed hierarchical solution that simultaneously meets the requirements of Industry 5.0.
[AI-44] When Is a Learned Command Adapter Worth It? Closed-Loop Identification and Counterfactual Auditing of Frozen Locomotion Policies
链接: https://arxiv.org/abs/2607.21867
作者: Zongtan Li
类目: Artificial Intelligence (cs.AI); Systems and Control (eess.SY)
备注:
Abstract:Adding a learned adapter to a frozen, command-conditioned locomotion policy is worthwhile only if the interface exposes improvements that are both real and recoverable from deployment-time observations. We introduce an adapter necessity audit that separates global operating-point gain,same-state counterfactual headroom, deployment gain over a cross-fitted fixed action, and state-allocation gain over a frequency-matched randomized policy. Source-cluster learner refits map these quantities and constraint violations to a GO/NO-GO/ABSTAIN decision. Closed-loop command- response identification provides optional decision features. On Go2, an archived scale-prefix diagnostic finds 5.2% same-state headroom but only 0.55% recovered allocation gain. Our confirmatory audit evaluates direct, scale, heading, and yaw interventions on twenty independent clusters for each of three query distributions induced by direct control, VGCC, and MPC, using 200 full learner refits. At 1% deployment and allocation thresholds and a 5% violation tolerance, direct queries return NO-GO, while VGCC and MPC queries ABSTAIN. VGCC has the largest mean deployment gain (1.34%), but its allocation lower bound is 0.09% and its violation upper bound is 6.25%. A deployment-representative twenty-cluster H1 audit also returns NO-GO, whereas a learner-level synthetic control returns GO. The audit therefore tests whether observable signal justifies state-dependent adaptation rather than presuming that an adapter is valuable.
[AI-45] DAGForge: Auditable Causal DAG Authoring with Biomedical Literature
链接: https://arxiv.org/abs/2607.21859
作者: Yi-han Sheu,Michael R. Steigman,Yu Zhou,Bo Wang,Fan-Yu Yen,Jordan W. Smoller
类目: Artificial Intelligence (cs.AI)
备注: 10 pages, 2 figures
Abstract:Constructing causal directed acyclic graphs (DAGs) is a core step in biomedical causal analysis, yet it remains a largely manual process. Analysts must connect study variables to prior literature, evaluate uncertain causal claims, and preserve sufficient provenance for expert review. We present DAGForge, a browser-based system for authoring causal DAGs as auditable, evidence-linked artifacts. Given free-text descriptions of study concepts, DAGForge creates a reproducible literature snapshot, uses an LLM-based reasoning module to generate structured pairwise causal judgments grounded in verbatim evidence excerpts, and assembles those judgments into a constraint-checked graph. Each proposed edge includes confidence estimates, provenance, and a reviewable rationale. The interface supports study specification, progress monitoring, evidence review, graph comparison, adjustment-set computation, and export. In evaluations against both compact benchmark DAGs and reference DAGs derived from published literature, DAGForge achieves high edge recall on the literature-based cohort while retaining verifiable evidence trails absent from LLM-only baselines. DAGForge thus reduces the burden of causal DAG curation while making the resulting assumptions auditable, supporting the design, analysis, and interpretation of biomedical studies.
[AI-46] oolGuardian: Declarative Security for AI Agent -Tool Interactions
链接: https://arxiv.org/abs/2607.21835
作者: Arun Ravindran,Saurabh Deochake
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI)
备注:
Abstract:LLM agents increasingly rely on external tools, expanding capability while creating a new security boundary: third-party tools may appear benign at the interface level while embedding unsafe behavior in implementation. Existing defenses rely on weak metadata, collapse characterization and policy judgment into a single decision, or use heuristic/LLM enforcement that lacks deterministic, auditable reasoning over task context and multi-tool composition. This paper presents ToolGuardian, a policy-driven framework for securing agent-tool interactions through pre-admission vetting and task-aware runtime authorization. ToolGuardian uses progressive characterization to convert evidence into structured facts: descriptions capture declared intent, system-call traces expose coarse behavior, mock execution reveals observed effects, and source analysis identifies latent behavior. ToolGuardian’s core contribution is an Answer Set Programming (ASP)-based declarative policy layer that reasons explicitly over capabilities, effects, task context, and composition. We compare ASP against heuristic and LLM-based policy realizations using identical inputs and output contracts. We evaluate ToolGuardian on 16 MCP-style tools, including 8 malicious variants derived from real open-source tools, and 20 runtime scenarios. For vetting, ASP reaches a deny-class F1 of 0.86 and 88% accuracy using description, syscall, and observed-effect evidence. For runtime authorization, fully specified realizations classify all scenarios correctly, while ablations show that removing compositional and conformance rules substantially degrades performance. Subjects: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI) Cite as: arXiv:2607.21835 [cs.CR] (or arXiv:2607.21835v1 [cs.CR] for this version) https://doi.org/10.48550/arXiv.2607.21835 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[AI-47] Probing Speaker Identity Sensitivity in Audio Deepfake Detectors
链接: https://arxiv.org/abs/2607.21820
作者: Daniyal Kabir Dar,Arun Ross
类目: ound (cs.SD); Artificial Intelligence (cs.AI); Cryptography and Security (cs.CR); Machine Learning (cs.LG)
备注: Accepted at IEEE/IAPR International Joint Conference on Biometrics (IJCB) 2026. 8 pages, 3 figures, 7 tables
Abstract:Audio deepfake detectors are trained to distinguish genuine speech from synthetic speech and often perform well on standard benchmarks. Yet the same detector that achieves less than 1% error on one dataset can see its error rate increase twentyfold when evaluated on a different dataset. We argue that one contributing factor is speaker-identity reliance: standard training corpora correlate speaker identity with the genuine/synthetic label, allowing detectors to partially rely on speaker-related cues rather than synthesis artifacts alone. We propose the Identity Sensitivity Score (ISS), a per-utterance diagnostic that quantifies how much a detector’s output changes across different speaker identity contexts. ISS requires no ground-truth labels at inference time and can be computed from the detector score and a pool of reference speaker examples. Across two detectors and two datasets, incorrectly classified utterances have ISS scores 29 to 52 times higher than correctly classified utterances, and ISS alone predicts misclassification with area-under-curve (AUC) up to 0.954. To test whether ISS actually captures identity-sensitive behavior rather than serving only as a proxy for prediction confidence, we apply voice conversion to 500 utterances and measure the resulting detector-score shift. Utterances flagged as identity-sensitive by ISS respond 19 to 30 times more strongly to this manipulation than utterances flagged as stable. These results position ISS as a practical inference-time diagnostic for speaker-dependent failure analysis in audio deepfake detection.
[AI-48] QLPO: Quadrant-weighted Sampling for Length-aware Policy Optimization
链接: https://arxiv.org/abs/2607.21793
作者: Siwei Chen,Siqi Chen,Xupeng Miao,Bin Cui
类目: Artificial Intelligence (cs.AI)
备注: 18 pages, 7 figures, 6 tables. Accepted at COLM 2026
Abstract:Recent large reasoning models often develop long chain-of-thought responses during reinforcement learning (RL), resulting in high inference latency and deployment cost. Existing methods for response length control typically rely on explicit length penalties or additional control modules, which require careful tuning and may compromise reasoning quality. We propose Quadrant-weighted Sampling for Length-aware Policy Optimization (QLPO), a simple resampling-based variant of GRPO that introduces implicit length control without modifying the reward function. QLPO first over-generates candidate responses and then resamples the training group by preserving the empirical correct/incorrect ratio while favoring short correct responses and long incorrect responses. This reshapes the training distribution and implicitly encourages shorter model outputs. Across models ranging from 1.5B to 32B parameters, including both base models and strong reasoning models, QLPO consistently improves the accuracy-length trade-off. It reduces response length by 30% to 70% while preserving reasoning performance. These results suggest that structured resampling provides an effective and robust approach to efficient reasoning.
[AI-49] MosaicJoin: Compact Semantic Sketches for Value-Level Join Discovery
链接: https://arxiv.org/abs/2607.21781
作者: Grace Fan,Eden Wu,Majid Daliri,Juliana Freire
类目: Databases (cs.DB); Artificial Intelligence (cs.AI)
备注:
Abstract:Join discovery is a core task in dataset search, enabling users to find columns that can be joined with a given query column. Early approaches focused on equi-joins, but data lakes and open-data repositories often contain columns whose values refer to the same entity but use different syntactic representations. To address this challenge, recent approaches discover semantically joinable columns but face a fundamental trade-off: methods that perform value-level comparisons accurately identify joinable columns but scale poorly to columns with high cardinality; column-level methods that encode an entire column into a single embedding are efficient but do not capture the fine-grained value alignment that determines whether a join is possible. We present MosaicJoin, a value-level semantic join discovery method that balances this trade-off. MosaicJoin achieves scalability through a novel sketching strategy that approximates the joinability of a column pair without having to compare all values. At query time, MosaicJoin scores each candidate sketch using a joinability score at a cost bounded by the sketch size, making retrieval efficient even for high-cardinality columns. A query subsampling operator further reduces online search time with provable accuracy guarantees, enabling robust retrieval for large query columns. Extensive experiments show that MosaicJoin outperforms previously published methods across all benchmarks while running up to 66 times faster than other value-level methods. MosaicJoin requires no training or fine-tuning, and it scales robustly to query columns containing up to 57K values and data lake columns containing up to 1M values.
[AI-50] Every Model Cheats: Prompt-Level Mitigation of Cheating on Offensive Cyber Tasks
链接: https://arxiv.org/abs/2607.21763
作者: Michael Kouremetis,Ads Dawson,Raja Sekhar Rao Dheekonda,Brian Greunke
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI)
备注: 21 pages, 4 figures, 7 tables
Abstract:Large language model (LLM) agents routinely cheat on cybersecurity benchmarks, inflating reported pass rates far beyond genuine capability. Prior audits of Cybench found cheating in 0.3-3.4% of traces, implicating only a handful of models. We present a controlled prompt-ablation study across 22 frontier models from 7 providers on 23 Cybench capture-the-flag (CTF) challenges under three prompt conditions (no anti-cheat, standard, severe). All 1,518 task traces were individually audited through a four-stage pipeline combining LLM-as-a-judge classification, programmatic verification, judge-verifier reconciliation, and human review. We find cheating is far more pervasive than previously estimated: under baseline conditions, 37.1% of passes involved cheating, 21 of 22 models cheated, and scores were inflated by up to 5x. Anti-cheat prompts reduce cheat propensity from 33.0% (baseline) to 17.8% (standard) to 8.5% (severe) without degrading, and sometimes improving, solve rates. However, even under the most restrictive prompt condition, eight models still produced cheated passes, four showed backfire effects, and cheating escalated from web search toward infrastructure probing. We introduce the “solve rate” metric (clean passes only) to distinguish genuine capability from cheated outcomes, and argue it should be standard practice in any evaluation where cheating vectors are available. Anti-cheat prompts are an effective and essentially free first layer of defense, but they are not a substitute for environmental controls.
[AI-51] What AI Red-Team Evaluations Can and Cannot Prove
链接: https://arxiv.org/abs/2607.21735
作者: Bandana Kaur
类目: Artificial Intelligence (cs.AI); Cryptography and Security (cs.CR)
备注: 21 pages, 4 figures, 5 tables. Code and data links are provided in the manuscript
Abstract:Red-team evaluations of AI models support some claims and not others, and the boundary between the two is calculable rather than merely a matter of judgment. We define the evidential ceiling of an evaluation as the largest factor by which one result can move belief under a fixed testing budget, derive it in closed form for the benchmark null result, and use it to locate that boundary exactly. We find that above a calculable harm rate, a benchmark of modest size certifies a category to a stated evidentiary standard, and a clean sheet is then the stronger of the two possible observations, outweighing a single reproduced failure. Below that rate, no passive benchmark of feasible size provides the specified evidence of safety under the fixed scoring rule and approximately independent trial structure. The crossing between the two regimes has a closed form. The bound is not specific to benchmarks: written in terms of a procedure’s hypothesis conditioned elicitation rates, it covers adaptive and automated red teaming as well, and shows that discrimination between the hypotheses rather than attack success is what determines evidential worth. Auditing eight evaluation suites against the boundary, we find that current benchmarks are adequate for high-frequency harm categories and several orders of magnitude short for rare, catastrophic ones. Safety benchmarks are not uninformative. They are informative about a specific and computable set of propositions, and the discipline they need is to state which.
[AI-52] A Defense of the Quadratic Model
链接: https://arxiv.org/abs/2607.21716
作者: Alexandru Meterez,Pranav Ajit Nair,Depen Morwani,Cengiz Pehlevan,Sham Kakade,Alex Damian
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Optimization and Control (math.OC); Machine Learning (stat.ML)
备注:
Abstract:Due to the complexity of neural network loss landscapes, optimization theory is forced to rely on idealized models, and there is generally a tradeoff between how theoretically tractable the model is, and how accurately it describes the true optimization dynamics. In this work, we stress test the simplest possible model of optimization – the quadratic model – and show that it can be surprisingly predictive in an LLM setting with 150M parameters and 3B training tokens. Specifically, we show that Taylor expanding the model and the loss function at intermediate checkpoints through training can accurately predict the optimization dynamics over windows that can last up to 10% of training. Having established this agreement, we then turn to analyzing the structure of these local quadratic optimization problems through two lenses: the Hessian spectrum and local stability. Using Lanczos quadrature with extremely deep probes, we are able to estimate the Hessian spectrum deep into the tail, and we find a surprising amount of structure in both the eigenvalues and eigenvectors, which depends on the batch size, preconditioner, and training time. We also empirically test local linear stability at intermediate checkpoints and compare it to theoretical predictions to demonstrate that optimization in LLMs typically occurs at a stochastic edge of stability, whose nature is also determined by batch size. Our results indicate the quadratic model may be a theoretically tractable proxy for pretraining optimization dynamics.
[AI-53] Persistent Computational State: A Session-Centric Runtime for Generative World Models
链接: https://arxiv.org/abs/2607.21686
作者: Zhen Lin
类目: Artificial Intelligence (cs.AI)
备注: 29 pages, 8 figures, 10 tables
Abstract:Generative world models are increasingly driven as simulators: a planner forks a state, rolls out futures, backtracks, and returns to a visited viewpoint. Recent benchmarks establish that current video world models fail this usage, and attribute it to the model, prescribing new architectures and training objectives. We show this attribution is incomplete, and for an important class of models simply wrong. Snapshotting the state the runtime already holds – an observation plus RNG state, a memory bank, or a windowed KV context, by architecture – and restoring it after a genuine excursion reproduces the never-left continuation byte-identically on all three; corrupting only the RNG degrades it. The capability was never missing: request-centric serving discarded it, inheriting from language-model serving the assumption that runtime state is recomputable – but world-model state carries a non-recomputable kernel. We define Persistent Computational State (PCS), the minimal non-recomputable state that must survive across requests, show it can be discovered by measurement, and build a session-centric runtime over it. Checkpoint and restore cost 0.012 ms against a 1.85 s generation step; resident sessions become host- rather than device-bounded (measured to 1,024); and world memory must be evicted by relevance to the return, not recency – the inverse of LLM practice.
[AI-54] Enhancing SLMs for Sustainable Code Optimization in Radio-Astronomy
链接: https://arxiv.org/abs/2607.21677
作者: Elisa Chiarotto,Jingbo Li,P. Chris Broekema,Rob V. van Nieuwpoort
类目: oftware Engineering (cs.SE); Artificial Intelligence (cs.AI)
备注:
Abstract:Recent Large Language Models (LLMs) can produce and optimize complex code. We investigate the use of LLMs to generate and optimize code for large-scale sciences, focusing on radio astronomy and sustainability. The LOFAR telescope is currently being upgraded, significantly increasing the sky area observed, while simultaneously processing more data faster. However, this is expected to increase the computational requirements 40-fold. This upgrade thus critically depends on rigorous performance optimization of existing software and widespread adoption of accelerators. The code base is very large, making this a daunting task. We therefore investigate and demonstrate an AI-driven approach meant to assist developers in evaluating and optimizing their code, including porting to hardware accelerators. The LOFAR community is committed to sustainable solutions, and needs to achieve these improvements without increasing the energy budget. We thus need to optimize existing codes or port them to accelerators, while making sure that the optimization process itself is also energy efficient. This poses a challenge, since LLMs are energy-intensive. We therefore propose to use Small Language Models (SLMs) instead to limit environmental impact. In this paper, we show how to enhance SLMs through the use of agentic AI. We extend the SLMs in two ways to improve code generation quality and performance: first with a multi-sampling generation strategy and second with incorporating compiler feedback. We demonstrate that multi-sampling SLMs can match or surpass larger single-generation models with fewer computational resources and that feeding compiler output back into the SLMs leads to consistent improvements across all tested models. Our approach is generic, and can also use Retrieval Augmented Generation (RAG) as well as static and dynamic analysis tools in the code generation pipeline.
[AI-55] Ordered Action Tokens for Visuomotor Policy Learning
链接: https://arxiv.org/abs/2607.21670
作者: Chaoqi Liu,Yue Zhao,Haonan Chen,Xiaoshen Han,Jiawei Gao,Ehsan Adeli,Yilun Du
类目: Robotics (cs.RO); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注:
Abstract:Action tokenization maps continuous robot action chunks to discrete tokens and has become an important interface for modern visuomotor policies. Existing approaches either rely on analytical discretization methods that produce prohibitively long token sequences or learned latent tokenizers that lack structure, limiting their compatibility with downstream policies. In this work, we identify three desiderata for action tokenization - high compression, total decodability, and an ordered token space - and introduce Ordered Action Tokenization (OAT), a learned action tokenizer that satisfies all three. OAT discretizes action chunks into an ordered sequence of tokens using a transformer with registers, finite scalar quantization, and ordering-inducing training mechanisms. By training each token prefix to decode into a valid action chunk, OAT places coarse control information in early tokens and uses later tokens to refine residual detail, yielding an anytime tradeoff between inference cost and action fidelity. We validate OAT in two prevailing uses of action tokens: autoregressive policies that generate tokens for control, and token co-training policies that use token losses to shape the vision-language model context consumed by a flow-based action expert. Across three policy backbones and more than 60 tasks spanning five simulation benchmarks and real-world settings, OAT consistently delivers strong policy performance while offering significantly greater flexibility at inference time.
[AI-56] Defining AI-Native Systems: Autonomy as Revision Authority
链接: https://arxiv.org/abs/2607.21659
作者: Cheng Tan
类目: Artificial Intelligence (cs.AI); Operating Systems (cs.OS)
备注:
Abstract:AI has begun to write systems code: agents now synthesize, verify, and deploy system components. Despite this shift, “AI-native” remains a marketing term with no precise technical definition. This paper gives it one. We define AI-nativeness along a single axis—authority over the system’s own decisions rather than by the capability of the underlying AI models. Building on a decision-level model of a system, we distinguish occupancy (who executes a decision) from revision authority (who may change it), organize revision authority into a ladder—self-tuning, self-rewriting, self-architecting and define a system as AI-native when an AI autonomously rewrites the system’s own implementations. The definition further requires an escalation detector, a verification procedure, and a verified fallback, while leaving purpose and correctness human-owned.
[AI-57] Cross-Model LLM Code Review: Should you use Claude to review Codex or vice versa? KDD’26
链接: https://arxiv.org/abs/2607.21656
作者: Zuodong Xiang,Yike Zhang,YueMing Zhang,Hailu Xu
类目: oftware Engineering (cs.SE); Artificial Intelligence (cs.AI)
备注: This paper had been accepted by Agentic SE @ KDD’26
Abstract:Developers increasingly use two coding agents together: one writes a draft, and the other reviews it. However, it is not clear whether the pairing is worth its cost and time, or whether the order of the pairing matters. We run a controlled experiment on 116 recent hard and medium lcb tasks with Claude and Codex across six conditions to approximate a software practitioner’s workflow: both solo baselines, both cross-model orderings, and both same-model orderings. The reviewer sees the problem and the writer’s draft but cannot execute tests, which approximates a code review step. Claude review raises Codex drafts from 71.6% to 89.7% ( p_BH=.001 ); Codex self review raises them to 84.5% ( p_BH=.022 ). The reverse direction does not pay off: Codex reviewing Claude drafts drops the pass rate from 91.4% to 82.8% ( p_BH=.046 ), and Claude self review leaves the 91.4% baseline unchanged. Our evaluation indicates that the useful pairing is asymmetric: use Claude to review Codex, not the other way around.
[AI-58] A Drift Stable Quantum Federated Learning for Intelligent Services
链接: https://arxiv.org/abs/2607.21647
作者: Shanika Iroshi Nanayakkara,Shiva Raj Pokhrel
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Quantum federated learning enables distributed clients to train quantum neural networks without sharing local data, making it promising for privacy-aware intelligent services. Intelligent services in this context refer to privacy-sensitive distributed decision systems, such as fraud detection and genomic classification, where reliable and fair client-level learning is as important as the accuracy of the aggregate model. However, heterogeneous client data and noisy quantum optimization often cause unstable local updates, client drift, and unfair performance between clients. This paper proposes DUQFL-Prox, a drift-stable quantum federated learning framework based on deep-unfolded local optimization. Instead of using a fixed local optimizer, each client performs adaptive unfolded SPSA updates, while a proximal term keeps the local model close to the global model. A lightweight controller learns step-specific optimization parameters to improve post-aggregation performance. Experiments on financial fraud and genomic classification tasks show that DUQFL-Prox improves stability, generalization, and client fairness compared with standard QFL baselines. The results suggest that deep-unfolded quantum federated learning can support more reliable and fair intelligent services in heterogeneous distributed environments.
[AI-59] Multi-Horizon Consistency as Geometry: When Latent Dynamics Contract and When They Do Not STOC
链接: https://arxiv.org/abs/2607.21645
作者: Kavya Bhand,Aadi Joshi
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注: 22 pages, 9 figures. Diagnostic study of multi-horizon latent consistency, expansion proxies (L20), and a stochastic-forcing law for world-model geometry. Code and experiment logs to be released publicly
Abstract:Multi-horizon latent consistency is a common training knob in video predictors and world models, but practitioners rarely know what it does to transition geometry. We treat lambda, the weight on multi-step latent agreement, as a diagnostic control and measure an empirical expansion proxy L20,q95 together with horizon-20 prediction error E20. On Moving-MNIST (n=6 seeds at the critical pair), raising lambda from 0 to 0.8 cuts L20 from 4.96 +/- 2.01 to 1.01 +/- 0.06 (paired t p=0.005, Wilcoxon p=0.031) and halves E20 (0.365 to 0.177, paired t p=1.1e-13). Four of six seeds cross L1 at lambda=0.8. The same loss does not produce population L1 on action-conditioned Pendulum-v1 or CartPole-v1, nor on KTH Actions video, even when E20 improves. An associational mediation analysis on MMNIST gives r-hat=0.94 (95% CI [0.88, 1.00], n=27, B=2000); lambda was not randomized. Defensive checks (architectural baselines, exogenous stress, WorldTest, MPC, scaling) mostly support a narrow claim: soft consistency can push passive video toward a near-contractive band, and that band is domain-limited. A stochastic-forcing law L20 ~ 1.23 + 1.82 eta at lambda=0.8 (bootstrap slope CI [1.73, 1.92], R^2=0.96) unifies control domains on the same curve via calibrated eta_eff. Complete joint slices at lambda in 0.4, 1.2 (30/30 cells, 5 eta x 3 seeds) show comparable linear L20(eta) slopes (~1.69 and ~2.00); we do not fit a continuous (lambda, eta) surface. We do not report DreamerV3 or TD-MPC2 returns.
[AI-60] ool-Guided Retrieval-Augmented Repair for Securing LLM -Generated C Code
链接: https://arxiv.org/abs/2607.21641
作者: Vidyut Sriram,Saatvik Pradhan,Suman Saha
类目: oftware Engineering (cs.SE); Artificial Intelligence (cs.AI)
备注:
Abstract:Large language models can generate C code from natural-language descriptions, but resulting programs often contain security vulnerabilities and compilation errors, posing risks for embedded and resource-constrained systems. This work investigates how feedback and retrieval improve reliability of LLM-generated C code. We present an analysis-and-repair workflow that combines compilation diagnostics, CodeQL static analysis, and KLEE symbolic execution with retrieval of prior repair patterns for iterative refinement. Evaluated on 5,000 C programming tasks exercising embedded relevant vulnerabilities, baseline models show substantial reliability gaps, with compilation failure rates up to 46% and security defect rates up to 49%. Our approach improves both metrics. For CodeLlama 7B, security defect rates decrease from 49% to 19% and total CodeQL errors drop from 15,088 to 2,463 (83.7%). For DeepSeek Coder 1.3B, compilation failures are reduced from 42% to 22% and security defects from 35% to 15%. These results show that integrating lightweight analysis tools can improve the safety of LLM-generated code for embedded development. Subjects: Software Engineering (cs.SE); Artificial Intelligence (cs.AI) Cite as: arXiv:2607.21641 [cs.SE] (or arXiv:2607.21641v1 [cs.SE] for this version) https://doi.org/10.48550/arXiv.2607.21641 Focus to learn more arXiv-issued DOI via DataCite
[AI-61] Measuring the Dependency Gap: Diagnosing Inter-Column Fidelity in Tabular Generative Models
链接: https://arxiv.org/abs/2607.21636
作者: Jie Zhang
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Synthetic tabular data is valued for preserving not only each column’s marginal distribution but the dependencies between columns – structure that carries much of the discriminative signal for minority classes in imbalanced domains such as fraud and clinical risk. Yet the metrics most commonly used to certify synthetic tabular data are, we show, largely blind to inter-column dependency: a baseline that models every column independently (and therefore destroys all dependency) is judged indistinguishable from real data by the logistic-regression C2ST, and the pairwise Trend score is only partially sensitive. We introduce a dependency-aware fidelity diagnostic that decomposes a strong classifier two-sample test (XGB-C2ST) into marginal, dependency, and numerical-categorical cross components, anchored between a worst-case fully-factorized reference (all dependency destroyed) and a best-case real-data oracle. Applying it to a state-of-the-art flow-matching generator (TabbyFlow/EF-VFM), we find a real dependency gap that standard metrics miss; destroying dependency outright collapses minority-class utility, and the generator’s residual gap carries a smaller, consistent utility cost. We then ask whether this gap reflects a structural limitation of mean-field generative objectives. It does not: consistent with recent recovery results for variational flow matching, the objective is asymptotically exact. Yet the gap is stubborn – a 16x increase in model capacity does not close it – pointing to the absence of direct dependency supervision rather than a capacity or structural limit. Consistent with this, and because the residual gap is higher-order, no cheap intervention closes it: not an in-model dependency mechanism, not post-hoc copula correction, and not the 16x capacity increase – a caution for a field that assumes such fixes help.
[AI-62] MotifRole-Diff: Risk-Optimal Role-Aware Corruption for Masked Molecular Graph Diffusion
链接: https://arxiv.org/abs/2607.21634
作者: Tasfia Nuzhat Ornee,Elias Hossain,Ivan Garibay,Niloofar Yousef
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Masked discrete diffusion for molecular graph generation typically applies a uniform corruption schedule to all tokens in a lossless graph-to-sequence representation, implicitly treating structurally heterogeneous molecular components as equally difficult and equally important to reconstruct. However, different molecular graph token roles exhibit substantial variation in denoising difficulty and their influence on the decoded molecule, motivating role-specific corruption strategies. We introduce MotifRole-Diff, a role-aware corruption process that allocates masking rates according to empirically measured denoising difficulty and graph-level perturbation impact while preserving the model architecture, clean sequence space, and lossless molecular-graph decoder. We formulate schedule selection as the risk-optimal allocation of a fixed masking budget across token roles. Our theorem characterizes optimality for the modeled role-weighted residual risk, while downstream generation performance is evaluated empirically. Under matched architecture, training budget, and sampling compute, MotifRole-Diff improves validity on QM9 from 0.905 to 0.944 while reducing FCD from 1.701 to 1.609, and on MOSES improves validity from 0.920 to 0.938 while reducing FCD from 2.125 to 1.850. Role-wise diagnostics further show improved reconstruction across molecular graph token categories. Together, these matched-compute results indicate that structurally informed corruption is a more effective masking strategy than uniform schedules for serialized molecular graph diffusion.
[AI-63] On the Depth Scalability of Logic Gate Networks
链接: https://arxiv.org/abs/2607.21633
作者: Taegun An,Dohun kim,Haebeom Lee,Changhee Joo
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Logic in Computer Science (cs.LO)
备注: 7 pages of main text, 4 figures, 2 tables
Abstract:Logic Gate Networks (LGNs) implement computation through compositions of Boolean operations, yet unlike classical Boolean circuits, existing LGNs do not reliably benefit from increased depth. We identify two distinct causes: optimization collapse in deep relaxed LGNs and a topology-induced limitation that persists even when skip-biased initialization and straight-through estimation stabilize training. Thus, trainability alone is insufficient; deeper layers must also receive information that supports useful computation. We introduce Input-Anchored Logic Gate Networks (IALGNs), in which each gate combines an evolving hidden feature with a direct input anchor. This topology preserves a computational spine while conditioning every layer on the original input. We show that a depth-D path can depend on up to D+1 input bits and establish a strict path-wise depth hierarchy. Random-k anchor relaxation further improves anchor selection without relaxing the spine. Across MNIST, CIFAR-10, and CIFAR-100, IALGNs achieve consistent fixed-width depth–accuracy improvements beyond 100 layers, whereas alternative LGN topologies saturate or degrade. Layer-wise probes, topology ablations, and effective-depth analysis show that input anchoring produces progressively more informative representations and preserves longer computational paths. These results demonstrate that scalable depth in LGNs requires both stable optimization and an information-access pattern that supports input-conditioned refinement. Comments: 7 pages of main text, 4 figures, 2 tables Subjects: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Logic in Computer Science (cs.LO) Cite as: arXiv:2607.21633 [cs.LG] (or arXiv:2607.21633v1 [cs.LG] for this version) https://doi.org/10.48550/arXiv.2607.21633 Focus to learn more arXiv-issued DOI via DataCite
[AI-64] Wavelet Phase Diffusion for Structurally and Semantically Consistent Sim-to-Real Translation
链接: https://arxiv.org/abs/2607.21628
作者: Kaiwen Wang,Frank Bieder,Yinzhe Shen,Carlos Fernandez,Jan-Hendrik Pauls,Omer Sahin Tas
类目: Artificial Intelligence (cs.AI); Image and Video Processing (eess.IV)
备注: Project Page: this https URL
Abstract:Simulation-to-reality translation must bridge the appearance gap between synthetic and real domains while preserving structural and semantic consistency. Conditioning-based methods achieve spatial alignment but introduce computationally expensive control modules. Paired-data methods achieve realism but rely on complex synthesis pipelines, often altering scene geometry and semantics. Training-free editing methods avoid both constraints but lack a learned appearance prior, limiting their perceptual quality. Recently proposed phase-preserving diffusion presents a promising alternative, but Fourier-domain formulations are constrained by global spectral coupling. This coupling induces spatial artifacts such as ringing and boundary leakage, thereby degrading structural and semantic consistency. We introduce Wavelet Phase Diffusion, which addresses this through two components. First, we operate in the Dual-Tree Complex Wavelet Packet Transform domain, whose localized wavelet packets enable spatially adaptive phase injection without global spectral interference. Second, Low-Frequency Randomization (LFR) replaces the low-frequency packet, decoupling the model from the synthetic illumination prior and enabling in-distribution real-world appearance. Both components train on unpaired open-domain data, and introduce negligible inference overhead. The spatial locality further enables instance-level translation, where individual objects or regions are translated to photorealistic appearance independently while the surrounding scene remains untranslated. On vKITTI \to KITTI image translation, ours outperforms prior methods in realism and semantic consistency while maintaining competitive structural alignment. For CARLA video translation, ours approaches the realism of paired-data methods while reducing VLM planner ADE and FDE by 5.4% and 5.1% , respectively.
[AI-65] Do Modules Stay in Their Lane? Role Drift in Compound LLM Systems
链接: https://arxiv.org/abs/2607.21627
作者: Xiaoyang Cao,Siddarth Srinivasan,Michiel A. Bakker
类目: Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注:
Abstract:End-to-end reinforcement learning can improve the accuracy of compound LLM systems, but it does not constrain how modules divide labor internally. We identify Role Drift, a failure mode in which modules preserve or improve end-task performance while deviating from their assigned roles through role-violating shortcuts that remain invisible to system-level evaluation. To make role drift observable and controllable, we propose Role Anchor, a regularizer that modulates how much each module deviates from its assigned role during end-to-end training. The key idea is to preserve how the role prompt shifts the module’s next-token predictions relative to a neutral prompt, which serves as a proxy for the role’s intended effect during training. Experiments on two compound LLM pipelines reveal role drift that accuracy alone fails to detect: a decomposer meant to split a question into sub-questions for a separate solver instead plants the answer in them, and a reader meant to answer from retrieved passages instead falls back on parametric memory. In fact, on the decomposer pipeline this shortcut drives most of the apparent RL gain: 86% of it vanishes once the decomposer is held to its role, indicating that terminal accuracy alone can badly overstate how much a compound system has genuinely learned. Across both pipelines, Role Anchor mitigates role drift at a tunable accuracy cost that varies by pipeline and anchor strength. Additional gradient analysis suggests that the regularizer reduces alignment with the role-drift direction rather than simply suppressing learning.
[AI-66] Discrete Action Space as a Prerequisite for GRPO Convergence in Small-Model Continuous Control
链接: https://arxiv.org/abs/2607.21626
作者: Dmytro Filatov,Valentyn Fedorov,Vira Filatova
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:We study whether Group Relative Policy Optimization (GRPO) can fine-tune small language models for simulated quadrotor continuous-control tasks. In our benchmark, vanilla GRPO fine-tuning of Qwen-0.5B for 25 Hz quadrotor velocity control collapses to the trivial zero action: 0 percent success rate, with entropy falling from 0.35 to 0.03 within 60 steps. Two ablations - removing the jerk-penalty term and removing the KL anchor to the pretrained prior - each prevent entropy collapse, yet neither enables learning. When the action interface is replaced by a 5-way categorical choice over PID presets, training converges. The resulting controller traces a smoothness-reliability Pareto frontier along training duration; both endpoints are reported: 98.6 percent success with 0.656 m/s3 jerk at 64 steps, and 100 percent success with 1.103 m/s3 jerk, or 0.796 under a matched velocity cap, at 256 steps. The recipe is evaluated across three pretrained language models. As context, a re-tuned classical baseline, PID with Ki = 0.30 and vmax = 2.5, reaches the same 100 percent success rate at jerk 0.736 m/s3. A high-fidelity simulation using Crazyflie 2.1 dynamics surfaces a hover-region training-distribution gap.
[AI-67] rajectory-Aware Retrieval Agents for Temporal Decision- Making
链接: https://arxiv.org/abs/2607.21625
作者: Jing Wang,Jie Shen,Xing Niu
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:We study the problem of decision-making from long-form, temporally structured text using large language model (LLM) agents. Standard retrievalaugmented generation (RAG) pipelines fragment chronological context into isolated snippets, discarding the temporal structure that is often critical for correct downstream decisions. We introduce TLM (Trajectory Language Model), a closed-loop agentic framework that iteratively refines the evidence set using SHAP-guided feedback. The key technical contribution is the latent growth curve model (LGCM) over retrieved chunk embeddings, which provides an interpretable mechanism for detecting trajectory trends, turning points, and information gaps. We show that, under a scorer-calibration assumption (which holds approximately in practice), the iterative refinement procedure is monotonically non-decreasing in the probability assigned to the correct label. Empirically, TLM is evaluated on three temporally grounded decision tasks: medical question answering, earnings call surprise prediction, and overnight stock gap prediction. TLM substantially outperforms both zero-shot LLM baselines and standard retrieval-augmented approaches on the medical task, and yields consistent, economically meaningful gains on the two financial tasks.
[AI-68] FBLayout: Optimizing Memory Layout for Efficient LLM Finetuning on Mobile GPUs
链接: https://arxiv.org/abs/2607.21624
作者: Kahou Tam,Wei Niu,Yu Bao,Xiaomin Ouyang,Chengzhong Xu,Li Li
类目: Artificial Intelligence (cs.AI); Distributed, Parallel, and Cluster Computing (cs.DC); Machine Learning (cs.LG)
备注: 13 pages, 21 figures, Mobisys 2026
Abstract:Transformer-based models have enabled unprecedented capabilities across language, vision, and multimodal tasks. On-device fine-tuning of transformer models offers a privacy-preserving path to personalized AI, yet remains inefficient on mobile GPUs due to severe memory constraints and frequent layout transformations in attention mechanism during training. Existing mobile training frameworks either use unified layouts for forward and backward passes – leading to fragmented memory access and poor GPU utilization during backpropagation – or rely on explicit layout conversions, which introduce significant transformation overhead. To overcome this, we propose FBLayout, a layout-aware framework that co-designs tensor organization with mobile GPU platforms. FBLayout introduces: (1) a unified R-Tile layout for multi-dimensional reductions across forward/backward passes; (2) tile-based index transformation to eliminate physical data movement; and (3) activation-guided layout selection to propagate efficient layouts globally. Evaluations on seven transformer models across different mobile phones (including ARM Mali and Qualcomm Adreno GPUs) show that FBLayout achieves 2.2-5.7x speedup over MNN, TFLite, and TVM, while significantly improving cache efficiency and reducing memory footprint, enabling practical on-device large model fine-tuning. Comments: 13 pages, 21 figures, Mobisys 2026 Subjects: Artificial Intelligence (cs.AI); Distributed, Parallel, and Cluster Computing (cs.DC); Machine Learning (cs.LG) Cite as: arXiv:2607.21624 [cs.AI] (or arXiv:2607.21624v1 [cs.AI] for this version) https://doi.org/10.48550/arXiv.2607.21624 Focus to learn more arXiv-issued DOI via DataCite (pending registration) Related DOI: https://doi.org/10.1145/3745756.3809214 Focus to learn more DOI(s) linking to related resources
[AI-69] From Profiles to Steering Vectors: Global Sparse Priors and Local Semantic Calibration for Personalized Text Generation
链接: https://arxiv.org/abs/2607.21620
作者: Liuji Chen,Zeyu Zhang,Xinyuan Zhang,Shuai Nie,Qiang Liu,Shu Wu,Liang Wang
类目: Artificial Intelligence (cs.AI)
备注: Under review
Abstract:Personalized text generation requires models to capture user-specific writing styles from historical data. Existing approaches based on retrieval, parameter-efficient fine-tuning, or activation steering either introduce inference and storage overhead or struggle to separate stylistic signals from semantic content. We propose GLASS, a training-free framework for personalized generation via Global-Local Activation Steering with Sparse priors. GLASS uses sparse autoencoders to extract a global user-style prior from historical responses and constructs local contrastive style vectors over clustered interaction scenarios. During inference, it jointly injects global and local vectors into different model layers, enabling context-aware personalization without retrieval or parameter updates. Experiments on LaMP and LongLaMP show that GLASS outperforms retrieval-, fine-tuning-, and steering-based baselines across ROUGE metrics and LLM-as-judge evaluations. Further analyses show that SAE-based representations are more robust to topic and length shifts, suggesting better disentanglement of stylistic information from semantic residue.
[AI-70] LeafData: An Agent ic System for Data Migration
链接: https://arxiv.org/abs/2607.21618
作者: Sadanand Katukuri,Rajasekhar Bada,Navya Induri,Rohit Gandham,Lynette Pinto,Joses Selvan,Abishek Krishnamoorthy,Joseph Rozario,Pu Tian,Pavan Poudel,Yalong Wu
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Modern data migration relies on JSON configuration to define data connection, pipeline logic, and orchestration behavior. This requires domain knowledge from users and is time-consuming and error-prone. In this paper, we present LeafData, an agentic system that converts user intent into validated and executable JSON configuration for data migration. Specifically, LeafData comprises a frontend chatbot and the backend service. The chatbot incrementally collects required information from users and performs schema-driven validation, while the backend service processes validated inputs and generates JSON configuration artifacts. These artifacts are directly consumable by orchestration platforms, enabling end-to-end pipeline generation and execution without manual coding. LeafData supports heterogeneous data migration across various data sources and connectors including relational databases, file-based systems, document-oriented databases, and REST APIs.
[AI-71] Lost in Context: Addressing Context Anxiety in Large Language Models ICML2026
链接: https://arxiv.org/abs/2607.21616
作者: Ifueko Igbinedion,Jillian Ross,Etienne Ricardez,Sertac Karaman,Eric So
类目: Artificial Intelligence (cs.AI)
备注: Accepted at ICML 2026. 17 pages
Abstract:Conventional wisdom suggests that reasoning models fail when problems exceed their capabilities. However, we find that frontier reasoning models sometimes possess the necessary capabilities to solve problems but fail due to premature self-doubt – a phenomenon informally known as context anxiety. We provide the first systematic study of context anxiety, demonstrating that it arises, in part, from a model’s inability to accurately estimate the tokens required to complete a task. We also show that context anxiety leads to material efficiency losses when models operate under perceived constraints. Building on this analysis, we further show that models can learn alternative strategies for solving long-horizon problems without exhibiting context anxiety, suggesting that performance improvements may be achievable not through scaling model capabilities, but by improving models’ ability to accurately assess and adapt to their own limitations.
[AI-72] FrED: External Data Influence Estimation via Domain Knowledge Graph Grounding
链接: https://arxiv.org/abs/2607.21615
作者: Theodoros Aivalis,Iraklis A. Klampanos,Antonis Troumpoukis,Joemon M. Jose
类目: Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注:
Abstract:The rapid deployment of generative AI has amplified the critical need for Training Data Attribution to ensure transparency and accountability. However, current parametric approaches require computationally prohibitive access to model weights, while similarity-based methods ignore deep structural context. We propose a novel probabilistic framework that operates entirely in a black-box setting. Our method fuses continuous feature similarities with discrete, domain-specific Knowledge Graphs (KGs). This approach ensures the attribution is grounded in structural reality, explicitly rewarding highly specific historical samples while preventing generic background data from dominating the results. We evaluate our framework across two distinct domains where linking outputs to data and domain context is inherently complex: abstract artistic image synthesis and high-dimensional physical weather forecasting. Extensive benchmarking demonstrates the robust efficacy of our approach. In the artistic domain, it achieves a strong Linear Datamodeling Score that exceeds standard black-box similarity baselines, while closing much of the gap to gradient-based estimators. We additionally present a cross-domain feasibility case study in environmental forecasting, where we use domain KGs to retrieve physically consistent historical analogs for regional flood forecasts, improving geographic localisation over a latent-only baseline. Operating entirely without internal model access, our approach provides an efficient, interpretable mechanism for post-hoc influence analysis and domain-grounded retrieval.
[AI-73] Household Movement Detection in Mixed-Format Occupancy Data Using LLM -Based Entity Resolution
链接: https://arxiv.org/abs/2607.21614
作者: Sasirekha Oguri,John R. Talburt,Mert Can Cakmak
类目: Artificial Intelligence (cs.AI)
备注: Computer Science Information Technology (CS IT) ISSN : 2231 - 5403, Volume 16, Number 10, May 2026
Abstract:Entity resolution (ER) typically relies on pairwise similarity comparisons between records, which limits its ability to capture indirect relationships present in demographic occupancy data. An important indirect pattern arises from household movement, where multiple individuals relocate together across addresses, but detecting such patterns is difficult due to mixed-format records, noise, duplication, and the absence of stable identifiers. This paper proposes an AI-enhanced framework for detecting indirect entity links associated with household movement in unstandardized name-address data. The approach integrates prompt-based large language model (LLM) named entity recognition for extracting personal names and addresses without extensive preprocessing, semantic text embeddings for robust similarity computation, and graph-based reasoning to infer group-level movement patterns. Experimental evaluation on SPX benchmark datasets (S8-S12) generated using the Synthetic Occupancy Generator demonstrates that incorporating indirect household movement evidence improves recall by 8-15% while maintaining high precision, yielding F1-score gains of 6-8% over a strong pairwise baseline.
[AI-74] Procedural Knowledge Is Not Low-Rank: Why LoRA Fails to Internalize Multi-Step Procedures
链接: https://arxiv.org/abs/2607.21612
作者: Simon Dennis,Kevin Shabahang,Hao Guo,Rivaan Patil
类目: Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: 14 pages
Abstract:Parameter-efficient fine-tuning methods like LoRA have become the default for adapting large language models, succeeding across instruction following, style transfer, and factual adaptation. We show that for procedural knowledge–the ability to follow multi-step procedures with conditional branching through to terminal states–LoRA fails to match full fine-tuning at the ranks where it retains its efficiency advantage. In a systematic ablation (r = 16–128) on a procedural travel booking task (14 nodes), all LoRA configurations fail uniformly (task success = 2.54 vs. 4.11 for full fine-tuning, all p 0.001), with scores decreasing at higher ranks–despite maintaining 95–99% conversation completion rates. Cross-domain replication on Zoom support (14 nodes) and insurance claims (55 nodes) at 8B confirms the failure generalizes: LoRA underperforms full fine-tuning by 0.8–2.2 points on average at both r = 32 and r = 128, with the largest gap on the most complex procedure. Quadrupling rank from 32 to 128 provides marginal improvement but does not close the gap. SVD analysis of the weight changes produced by full fine-tuning explains why: across three domains at both 3B and 8B, the mean effective rank of the update ranges from 761 to 1,026, and rank 128 captures only 43–51% of the squared Frobenius norm. Together, these findings establish that for procedural tasks LoRA falls well short of full fine-tuning–a fundamental limitation for agentic applications.
[AI-75] SCOPE and SCION: A Benchmark and an Auditable Reference Pipeline for Schema Induction and Fusion from Text
链接: https://arxiv.org/abs/2607.21610
作者: Miaobo Hu,Xiaobo Guo,Shuhao Hu,Bokun Wang,Rui Chen,Xin Wang,Daren Zha,Jun Xiao
类目: Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注:
Abstract:Schema graphs are an upstream bottleneck of schema-grounded information extraction and knowledge graph construction, yet most extraction systems assume the schema is already available. We introduce SCOPE (Schema Construction and Ontology-induction Pipeline Evaluation), a train-text-only benchmark for corpus-to-schema induction and optional schema fusion from raw text, built from 24 public information extraction sources (15 RE and 9 EE) normalized into evaluation-only gold schema graphs; its core event-extraction target covers event types and within-event argument roles, with inter-event links reported separately. We present SCION (Schema Construction and Induction with Ontology Normalization), an auditable reference pipeline rather than a new extraction architecture; it constructs candidate spaces from train text and restricts naming, merging, filtering, validation, and conservative fusion to candidate-linked evidence under strict JSON contracts. On the SCOPE core suite, SCION-lite attains the highest F1 among released source-schema references, Text2Onto-style, LLM-only, and matched extract-then-aggregate baselines under Literal, Fuzzy, Continuous, and Graph schema-graph metrics, while the compact open-model SCION-RL variant reduces reliance on proprietary LLM schema engineers. These results are reported against normalized typed-edge targets rather than as claims that induced schemas surpass human ontology design; the release includes evidence-linked outputs, parse/fallback logs, candidate retention/merging logs, run manifests, code, and benchmark packages at this https URL.
[AI-76] Coupled Hierarchical Search over Topology and Execution for Agent ic Workflow Synthesis
链接: https://arxiv.org/abs/2607.21609
作者: Dong Li,Yanchi Liu,Xujiang Zhao,Wei Cheng,Zhengzhang Chen,Xintao Wu,Zhong Chen,Chen Zhao,Haifeng Chen
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Although structured workflows empower Large Language Models (LLMs) to tackle complex problems, automating their creation is severely hindered by a vast combinatorial search space, frequently resulting in inflexible and resource-heavy offline training dependencies. To address this, we conceptualize workflow generation as an intertwined topology-and-execution search paradigm, where the broader topological layer dictates subtask boundaries and lower-level execution outcomes actively reshape the topology itself. Building on this foundation, we introduce HierFlow, a training-free, test-time hierarchical search architecture that automates agentic workflow design by merging feedback-guided topology adjustments with a fast, MCTS-inspired tree search for sub-workflow optimization. HierFlow maximizes efficiency through an intelligent gating module that selectively triggers execution-level searches based on contextual necessity, a mechanism we further support with an in-depth analysis detailing how varying degrees of cross-task coupling impact the effectiveness of hierarchical splitting. Comprehensive testing across question answering, mathematical reasoning, and code generation benchmarks confirms that HierFlow consistently outperforms strong baselines, delivering an optimal balance of high-quality results and computational efficiency without any additional training overhead.
[AI-77] Spectral Flow Certificates for Depth-Aware Long-Range Propagation in Graph Neural Networks
链接: https://arxiv.org/abs/2607.21607
作者: Ranjan Veerabhadraswamy,Ajith Jubilson Emerson
类目: Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: 16 pages, 15 figures
Abstract:Graph Neural Networks propagate information through local message passing, but the graph topologies themselves can silently prevent any amount of training from solving long-range tasks. When we deploy GNNs on new graphs, there is currently no inexpensive way to know, before training begins, whether the graphs’ structures will allow information to travel far enough between distant nodes. We address this gap by proposing Spectral Flow Certificates (SFCs), single scalars computed from the graphs’ normalised Laplacians in seconds, requiring no model training and no labelled data. An SFC fuses a graph’s algebraic connectivity with the chosen message-passing depth into one number that measures how much of the critical spectral bottleneck can be traversed within the available depth budget. Unlike raw spectral gaps, which are static and depth-agnostic, SFCs adapt as the number of layers increases and therefore carry strictly more diagnostic information when depths vary. Compared with classical structural statistics such as average effective resistance and graph diameter, SFCs explain more than twice as much variance in trained GNN long-range accuracy. Across twenty-five synthetic graph families spanning paths, cycles, grids, regular graphs, and random graphs, SFCs predict trained accuracy before any gradients are computed, achieving explanatory power above ninety percent at all tested depths. The same predictive relationships hold on one hundred fifty real molecular graph topologies drawn from three independent benchmark datasets, confirming that the findings are not artefacts of their synthetic construction. Taken together, these results show that a single eigenvalue computation is sufficient to flag topology-limited graphs before committing to expensive training pipelines, providing a principled first filter for GNN deployments.
[AI-78] ILT: Improving Compositional Generation in Diffusion Models with a Model-Intrinsic Reward
链接: https://arxiv.org/abs/2607.21606
作者: Debottam Dutta,Jaehoon Hahm,Jianchong Chen,Romit Roy Choudhury
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Recent advances in powerful text-to-image generation models have made it increasingly important to develop test-time methods that modify the sampling trajectory to produce images more faithful to complex compositional prompts. We present TILT, a training-free framework for compositional text-to-image generation via test-time reward alignment. We interpret compositional failures as overlap modes between joint and single-concept distributions, and define a reward that favors samples where all concepts are jointly present. This reward is intrinsic to the base model and does not require any external supervision or reward models. This yields a KL-constrained objective with a closed-form tilted target distribution and principled guiding steps for diffusion sampling. The interaction of concept distributions together with the above reward naturally leads to two different guidance strategies while a hybrid approach that balances their respective benefits produces stronger performance. Experiments on prompts from T2ICompBench show that our method improves compositional alignment while preserving image quality compared to previous baselines.
[AI-79] Agent KVShift: Efficient KV Cache Reuse for Agent ic Memory Systems
链接: https://arxiv.org/abs/2607.21604
作者: Nilesh Prasad Pandey,Jason Kong,Lanxiang Hu,Quanling Zhao,Yujie Zhao,Onat Gungor,Hao Zhang,Tajana Rosing
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Memory-augmented LLM agents maintain context across hundreds of interactions through agentic memory systems that actively curate retrieved content with LLM-generated metadata such as summaries, keywords, and tags. From an inference cost standpoint, every retrieval triggers a full re-encoding of these structured memory units into Key-Value (KV) states, which dominates prefill latency. Existing training-free KV reuse methods mitigate this by selectively recomputing a small fraction of tokens, but were designed for RAG-style raw passages and degrade on structured agentic memories. We present AgentKVShift, a training-free, probe-guided KV residual correction method that operates per retrieved memory unit. One of the crucial insights we demonstrate is that the per-memory KV reuse residual decomposes into a shared memory-level offset plus small token-wise fluctuations. Estimating this offset from a small probe set allows us to correct every reused token by a single weighted correction. Unlike prior reuse methods which decide which tokens to recompute and leave the rest of the cache stale, AgentKVShift also corrects the tokens it does not recompute, turning the refresh budget into useful signal across the entire chunk. Across four open-source LLMs spanning 3B to 32B parameters and two long-horizon agentic memory benchmarks (long-term dialogue and agentic applications), AgentKVShift achieves near full recompute performance while refreshing only 10-30% of the cache, outperforming baselines at the same recompute ratio. It requires up to 5x lower recompute to reach this near-full performance, which prior reuse methods only attain at 45-55% refresh. In this regime, AgentKVShift delivers prefill speedups of 2-3.5x over no-KV-reuse on a single A100. AgentKVShift orthogonally composes with KV cache quantization, retaining over 2x the F1 of prior reuse methods under aggressive 2- and 4-bit settings.
[AI-80] ransferable Latency Prediction for Fast LLM Screening on Heterogeneous Edge Devices
链接: https://arxiv.org/abs/2607.21602
作者: Xiaolong Tu,Vinod K. Mishra,Venkat R. Dasari,Anu G. Bourgeois,Haoxin Wang
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Accurate latency prediction is critical for deploying large language models (LLMs) on heterogeneous edge devices, where inference latency is affected by model architecture, prompt behavior, runtime backend, hardware utilization, dynamic voltage and frequency scaling (DVFS), and thermal variation. This paper presents a runtime-aware latency prediction framework for deployment-oriented LLM selection. The framework represents each inference request as a hardware-runtime-model-prompt configuration, separates inference into prefill and decode phases, and adaptively fuses static descriptors with dynamic hardware telemetry through a gated prediction model. We evaluate the framework using Pixel mobile devices and validate the profiling pipeline on Jetson Nano, Orange Pi 5 Pro, and an RTX 3090-class GPU platform. On Pixel 8, the full predictor improves total-latency R-squared from 0.953 to 0.960 and decode-latency R-squared from 0.957 to 0.973 over a static-only baseline. On Pixel 8 Pro, it improves prefill-latency R-squared from -1.383 to 0.966. For cross-device transfer, calibration improves Pixel 8 Pro to Pixel 8 total-latency R-squared from -0.974 to 0.940 and decode-latency R-squared from -1.085 to 0.927. Heterogeneous profiling further shows that latency is highly device- and runtime-dependent: the same SmolLM2 model family reaches 8.42 tokens/s on Orange Pi 5 Pro but 64.38 tokens/s on an RTX 3090-class GPU. These results demonstrate that runtime-aware prediction with lightweight calibration can reduce profiling cost and support latency-aware LLM deployment across heterogeneous edge platforms. Subjects: Artificial Intelligence (cs.AI) Cite as: arXiv:2607.21602 [cs.AI] (or arXiv:2607.21602v1 [cs.AI] for this version) https://doi.org/10.48550/arXiv.2607.21602 Focus to learn more arXiv-issued DOI via DataCite
[AI-81] From Frame-Level Recognition to Event-Level Confirmation: Repair Traces and Runtime Failure Analysis of Public-Space Gesture Interaction
链接: https://arxiv.org/abs/2607.21601
作者: M. Meng,Yansong Zhang
类目: Artificial Intelligence (cs.AI)
备注: 15 pages, 5 figures, 10 tables
Abstract:Public-space gesture interaction is often evaluated as a frame-level recognition problem, but deployed systems expose a different failure boundary. In scenic kiosks, exhibition halls, and service terminals, users experience whether an intended action becomes a stable interaction event, not whether individual hand-landmark frames are correct. We call this the recognition-to-interaction gap. This paper analyzes 8 engineering repair records from a scenic-area interactive kiosk project, covering 4 gesture tasks: two-hand bowing, single-hand fist shaking, two-hand catching control, and knowledge-graph node hovering. From these traces, we extract 20 failure instances and organize them into six non-exclusive working failure classes: model-output degeneration, temporal mismatch, geometric-scale instability, coordinate-rendering mismatch, runtime lifecycle failure, and feedback synchronization and recovery failure. We further organize recurring repair mechanisms into an event-level runtime abstraction between the hand-landmark model and the interaction task. The contribution is deliberately bounded: a deployment-grounded failure taxonomy, an event-confirmation runtime abstraction, and case-study findings. We do not claim a new recognition model, large-scale user evaluation, or quantified accuracy gains. Comments: 15 pages, 5 figures, 10 tables Subjects: Artificial Intelligence (cs.AI) Cite as: arXiv:2607.21601 [cs.AI] (or arXiv:2607.21601v1 [cs.AI] for this version) https://doi.org/10.48550/arXiv.2607.21601 Focus to learn more arXiv-issued DOI via DataCite
[AI-82] Securing Multimodal AI through Internal Information Decomposition ICML2026
链接: https://arxiv.org/abs/2607.21600
作者: Jehyeok Yeon,Hyeonjeong Ha,Qiusi Zhan,Heng Ji
类目: Artificial Intelligence (cs.AI)
备注: Accepted as Spotlight Paper at ICML 2026
Abstract:Multimodal large language models introduce attack surfaces absent in unimodal systems: adversaries can distribute malicious intent across modalities to evade unimodal safeguards. This motivates using cross-modal consistency as a detection signal rather than inspecting each modality in isolation. Our key observation is that benign inputs induce compatible predictive behavior from text-only and vision-only reasoning that stabilizes when fused, whereas adversarial manipulation disrupts this consistency, causing abnormal multimodal behavior. Existing defenses that examine raw inputs or outputs overlook this internal fusion process, rendering them brittle and computationally expensive. We propose FlowGuard, a lightweight inference-time framework that detects harmful inputs by monitoring internal multimodal consistency. Unlike approaches that rely on scalar confidence metrics, FlowGuard derives FlowVectors inspired by Partial Information Decomposition that quantify cross-modal redundancy, synergy, and modality-specific dominance, capturing whether fused multimodal predictions remain aligned with unimodal semantic evidence. In a one-class classification problem trained solely on benign data, FlowGuard reduces Attack Success Rates from 90% to 15% on unseen attacks, with 3% utility loss and up to a 6 times latency reduction. Our results demonstrate that monitoring cross-modal consistency offers an efficient and effective defense for multimodal reasoning.
[AI-83] Decoupled Attention Fusion: Accelerating RAG with Efficient KV Cache Reuse EUROSYS2026
链接: https://arxiv.org/abs/2607.21599
作者: Xiabao Wu,Wentao Liu,Yongchao Liu,Jiajun Zheng
类目: Performance (cs.PF); Artificial Intelligence (cs.AI)
备注: Accepted by EuroSys 2026 (poster track)
Abstract:Retrieval-Augmented Generation (RAG) effectively mitigates hallucinations in Large Language Models (LLMs) but suffers from prohibitive Time-To-First-Token (TTFT) latency in long-context scenarios. Reusing pre-computed document KV caches addresses this but introduces a distribution mismatch, where offline caches lack the inter-document attention patterns required for coherent reasoning. CacheBlend reduces recomputation via selective attention, but suffers severe accuracy degradation at longer contexts. To address these challenges, we propose Decoupled Attention Fusion (DAF), a framework that maintains high accuracy while significantly reducing recomputation overhead. DAF decouples the attention process into three integrated stages: important-token self-attention to restore missing inter-document attention, question-document self-attention for standard inference, and a state fusion that concatenates their outputs to synthesize the final hidden states. By decoupling these operations into dense patterns, DAF is natively compatible with Flash-Attention kernels, maximizing hardware utilization without requiring complex attention masks. Experiments show that DAF delivers up to 2 times speedup over CacheBlend and 5.6 times over full recomputation with vLLM on long-context benchmarks, without sacrificing accuracy.
[AI-84] Risk Is Not the Target: A Monotonic Framework for Evaluating Wildfire Operational Risk Signals
链接: https://arxiv.org/abs/2607.21597
作者: Nicolas Caron,Christophe Guyeux,Maxime Coulmeau,Benjamin Aynes
类目: Artificial Intelligence (cs.AI)
备注: Accepted in 2026 IEEE 50th Annual Computers, Software, and Applications Conference (COMPSAC)
Abstract:Evaluating wildfire risk systems using standard machine-learning metrics such as F1-score or IoU is fundamentally flawed: these metrics assess event prediction accuracy, not the operational coherence of a continuous risk signal. This work proposes a novel monotonic evaluation framework that measures whether increases in a predicted risk score consistently correspond to increases in observed operational load, such as number of fires, intervention time, and deployed resources. Moreover, we compare three structurally different approaches on the French Alpes-Maritimes department: the expert-based DFE index, GRU- based predictive models, and FARS, a hybrid multi-agent system combining predictive AI with LLM-based reasoning. Experimental results reveal that the DFE, despite poor classification metrics, exhibits the most balanced monotonic behavior across the full risk scale. GRU models achieve strong local monotonicity but fail to produce well-distributed risk levels. FARS inherits and reveals the structural limitations of upstream signals rather than correcting them. The central finding is a paradigm shift: a good risk model does not predict fires accurately, but one whose ordinal scale meaningfully explains operational dynamics, as proved in this paper. Code of the monotonic framework is available on github.
[AI-85] FlowEvo: Self-Evolving Agents through the Co-Evolution of Workflows and Executable Skills
链接: https://arxiv.org/abs/2607.21596
作者: Zeyu Ren,Ling Yue,Ran Li,Yishu Wang,Shengxiang Xu,Hanmo Liu,Shaowu Pan,Shimin Di
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Large language model agents increasingly solve complex tasks by constructing inference-time workflows that combine reasoning, tool use, and code execution. While such workflows enable flexible problem solving, the useful procedures discovered during execution are often transient: they help solve the current task but are not retained in a form that can systematically benefit future tasks. We present FlowEvo, a training-free framework that compiles successful traces into reusable skill records. Each record pairs a callable artifact with auxiliary structured guidance, and admission applies interface, replay, and safety checks where feasible. These skill records persist in a skill bank at inference time. FlowEvo is organized around three coupled mechanisms: (1)~workflow-to-skill compilation, which extracts reusable executable artifacts from successful traces; (2)~skill-to-workflow feedback, which retrieves accumulated skills to support future problem solving through either direct execution or structured context injection; and (3)~skill curation, which monitors downstream utility and suppresses skills that cause negative transfer. Through this workflow–skill–workflow feedback loop, FlowEvo enables agents to accumulate and refine task-solving capability over time without updating model parameters. Experiments on benchmarks spanning interactive environments (ALFWorld) and code/math generation (HumanEval, GSM8K) show that FlowEvo achieves the best accuracy-cost tradeoff among the evaluated baselines under our implementation settings. On ALFWorld, FlowEvo achieves an 82.8% success rate, 23.6 percentage points above the strongest baseline, while its average token usage per episode is less than half that of the most efficient baseline. Controlled ablations confirm that each mechanism contributes to the overall result. The code is public at this https URL.
[AI-86] he Human-AI Substitution Principle: When will you be replaced by AI in your organization?
链接: https://arxiv.org/abs/2607.20781
作者: Bonny Banerjee,Shreya Singh
类目: Artificial Intelligence (cs.AI); Computers and Society (cs.CY); General Economics (econ.GN)
备注:
Abstract:Artificial Intelligence (AI) is rapidly transforming organizations, raising a fundamental organizational and economic question: when will a human employee be replaced by AI? We present an analytical model for studying Human–AI Task Allocation (HAT) in hierarchical organizations. A central feature of the HAT model is that it formally encodes the economic asymmetry between human skill acquisition and AI capability scaling. The HAT model allows us to derive how risk-adjusted costs, skills, organizational depth, deployment scale, strategic adaptation, and risk jointly determine when, where, why, and under what structural conditions human–AI replacement occurs. A key result is the Human–AI Substitution Principle, which provides a precise condition — grounded in the formal asymmetry assumption — under which AI replaces human labor. Building on this result, we show that AI adoption can produce abrupt workforce transitions, hybrid human–AI organizations, including cases where risk heterogeneity sustains human and AI roles without requiring a minimum-human-fraction constraint, and flatter managerial hierarchies with wider spans of control. The HAT model identifies structural conditions under which middle-management roles exhibit elevated vulnerability to automation, and shows that the vulnerability of highly skilled workers depends on a skill threshold shaped by organizational depth, baseline costs, and risk differentials. More broadly, the paper connects automation economics, organizational design, AI governance, and workforce planning into a unified theory of AI-driven organizational transformation.
[AI-87] AI-Driven Surrogate Models for Predicting Electrode-Scale Discharge Behavior in Lithium-Ion Batteries
链接: https://arxiv.org/abs/2607.20577
作者: Mengda Xing(CRIL, UA),Jean-Marie Lagniez(CRIL, UA),Alejandro Franco(LRCS)
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Physics-based simulations are essential for understanding the electrode-scale discharge behavior of lithium-ion batteries (LIBs) but suffer from prohibitive computational costs. To address this, we introduce a novel deep learning surrogate pipeline based on the Swin3D Transformer to predict spatiotemporal discharge dynamics directly from volumetric data. Our approach integrates two key innovations: Gaussian Positional Encoding (GPE), which enhances spatial feature representation by adapting to the complex geometry of electrode microstructures, and a specialized Temporal Encoding module to capture non-linear timeseries evolution. Experimental validation on an Electrochemical Simulation (ES) dataset demonstrates that our pipeline significantly outperforms state-of-the-art point cloud baselines in prediction accuracy. Furthermore, the proposed method reduces the computational overhead by orders of magnitude, providing a scalable and efficient framework for high-throughput battery design and optimization.
[AI-88] Ghost Vectors: Soft-Deleted Embeddings Remain Reconstructible in HNSW Vector Databases
链接: https://arxiv.org/abs/2606.18497
作者: Chandranil Chakraborttii,Jackeline García Alvarado,Sitora Abdulofizova,Shivanshu Dwivedi
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI); Databases (cs.DB)
备注: 13 pages, 5 figures, 12 tables. Prepared for submission
Abstract:Retrieval-augmented generation (RAG) allows large language models to access external and private corpora for factual, domain-specific responses. Modern RAG pipelines use hierarchical navigable small world (HNSW) vector databases for efficient similarity search. When a user requests data deletion, the systems typically only mark the record as deleted, leaving the embedding on disk physically unchanged. This soft-delete operation raises compliance concerns under data-erasure and retention requirements such as GDPR Article 17 and HIPAA. Analysis on three HNSW implementations confirms that deleted vectors remain physically recoverable by accessing the raw index files at the storage layer, bypassing API access. Using the Vec2Text inversion model without domain-specific fine-tuning, we show this vulnerability on multiple real-world datasets and data modalities. On Wikipedia biographical living persons dataset (BLP), we successfully recover 25.5% of exact person names and 46.4% of geographic locations (ROUGE-L 0.185). Recovery reaches 100% for both patient age and gender markers (ROUGE-L 0.290) on highly structured, sensitive data (NIH Synthea dataset). On soft-deleted image embeddings, we show 100% tissue classification on histopathology patches (p=1.02e-07) and top-1 identity recovery reaches 99% on facial embeddings (p0.01). This work introduces Epoch Key Rotation, which encrypts vectors and discards the key upon deletion. Epoch key rotation reduces observed PII recovery to 0% and completes in 2.5 ms for 500 deleted vectors (approximately 0.005 ms/record). Additionally, it generates an ECDSA-signed cryptographic proof as an auditable record of the deletion event.
[AI-89] Quantum Spectral Model: Data Reuploading with Input-Conditioned Frequency Support
链接: https://arxiv.org/abs/2607.22516
作者: Peiyong Wang,Udaya Parampalli,Casey R. Myers
类目: Quantum Physics (quant-ph); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: 53 pages, many figures
Abstract:A central design principle in modern machine learning and artificial intelligence is to align a model’s inductive bias with the structure of its input data. For matrix-valued inputs, relevant matrix-level relationships can be characterised through spectral values and spectral subspaces; however, common coordinate-wise rotation-gate data-encoding unitaries used in most quantum machine learning models do not explicitly construct such a matrix-level representation. We introduce Quantum Spectral Models (QSMs), in which we construct the generator of the data-encoding unitary directly from each input matrix. We study three QSM variants based on symmetric, global block, and non-overlapping patch-local block Hamiltonians. Their outputs admit truncated Fourier representations in which input-dependent spectral gaps supply candidate phase carriers, while spectral subspaces help determine their coefficients. We evaluate the QSMs and comparison quantum models on two matrix representations of Pendigits and two controlled synthetic tasks defined by spectral statistics. At the largest evaluated circuit depth, QSM variants lead the tested quantum models in mean test accuracy across all four benchmarks. The patch-local QSM leads on Pendigits, whereas the global block-Hamiltonian QSM leads on the controlled spectral tasks. Ablations show a task-dependent reversal: subspace-preserving controls perform better on Pendigits, whereas spectral-value-only controls lead among the tested ablations on the synthetic tasks. Together, these results shed new light on quantum machine-learning model design by showing how input-conditioned spectral representations can provide an analysable inductive bias, while offering a broader perspective on structure-aware model design in machine learning and artificial intelligence.
[AI-90] CausalForge: A Formally Grounded Self-Improving Agent ic Framework for Automated Research in Causal Inference
链接: https://arxiv.org/abs/2607.22511
作者: Jiyuan Tan,Vasilis Syrgkanis
类目: Machine Learning (stat.ML); Artificial Intelligence (cs.AI); Machine Learning (cs.LG); Econometrics (econ.EM)
备注:
Abstract:Automating theoretical research is constrained not only by the generation of candidate results, but also by their reliable evaluation. A common approach is to close the research loop with a large language model (LLM) reviewer. However, such reviewers remain empirically unreliable: they may accept fabricated papers and detect them at rates close to chance (Bad Scientist, 2025). We present CausalForge, a framework for automated theoretical research in causal inference grounded in the Lean proof assistant. CausalForge combines Causalean, a foundational Lean library for causal inference containing 7,035 machine-checked declarations developed with language-model assistance under human design and review, with CausalSmith, a self-improving agentic pipeline that selects research topics, proposes results, formalizes statements, constructs proofs, and presents the resulting artifacts for human inspection. Because a machine-checked proof establishes only that a formal statement follows from its assumptions, not that the statement faithfully captures the intended scientific claim, the pipeline augments kernel verification with a statement audit that compares each formal theorem against the informal claim it is intended to express. We evaluate the system using artifacts produced by completed autonomous research runs. The source code, formal library, and run records are available at this https URL.
[AI-91] Learning to Prepare Molecular Ground States with Transformer Models
链接: https://arxiv.org/abs/2607.22468
作者: Alex Koziell-Pipe,Jasmine Brewer,Jem Guhit,Marwa H. Farag,Kripa Panchagnula,Gabriel Laude,Fabian Finger,Carlo Gaggioli,Ludmila Szulakowska,Oliver J. Backhouse,Christos Papalitsas,Jason G. Mustakis,Thomas Soini,David Munoz Ramo,Stephen Clark,Elica Kyoseva,Enrico Rinaldi
类目: Quantum Physics (quant-ph); Artificial Intelligence (cs.AI)
备注:
Abstract:Quantum state preparation is a key component of many quantum algorithms. Performing this step efficiently is essential for realizing practical quantum advantage in quantum chemistry applications. Iterative algorithms like ADAPT-VQE can produce shallow ground-state preparation circuits, but become computationally prohibitive for the larger molecules relevant to materials science and pharmaceutical development. Here, we introduce ADAPT-GQE, a generative AI framework that learns to synthesize ground-state preparation circuits for electronic structure calculations. We first use ADAPT-VQE to generate high-quality reference circuits, which are then used as targets for training models for circuit generation. Once trained, the model can efficiently propose and score circuits, enabling reinforcement learning (RL) to drive circuit generation accuracy beyond the accuracy of the ADAPT-VQE training data. This pipeline achieves order-of-magnitude reductions in circuit generation time relative to ADAPT-VQE while maintaining comparable or improved state-preparation accuracy. We demonstrate ADAPT-GQE on imipramine, a well-established tricyclic antidepressant that serves as a representative, challenging target for computational modelling in drug stability protocols. We execute generated circuits on Quantinuum Helios-1, representing a milestone for AI-generated quantum chemistry circuits on state-of-the-art quantum hardware. These results establish a pathway toward automated quantum circuit synthesis for utility-scale quantum computational chemistry.
[AI-92] PRIMS: Physics-guided Representation for Fluid Identification in Multimodal Sensing
链接: https://arxiv.org/abs/2607.22422
作者: Hai-Long Nguyen,Trung Thanh Nguyen,Lars Holm,Dennis Alveringh,Duc Viet Le
类目: Fluid Dynamics (physics.flu-dyn); Artificial Intelligence (cs.AI)
备注: 2026 European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases
Abstract:Accurate on-device fluid identification is essential for microfluidic applications, yet maintaining reliability under varying flow, pressure, and temperature remains a key challenge. Existing learning-based methods often treat sensor signals as domain-agnostic features, neglecting the underlying physical relationships that govern fluid behavior, thereby limiting generalization and interpretability. To address this, we propose PRIMS, a physics-aware multimodal Transformer that integrates physical knowledge into representation learning and attention mechanisms through three dedicated modules: (1) Physics-based Token Vectorization transforms raw Coriolis and pressure sensor signals into physically meaningful token embeddings; (2) Physical Component Synthesizer models viscosity-related dependencies among flow, pressure, and density; and (3) Physics-guided Fusion captures cross-physical correlations through attention-based integration. By embedding these physics-based relationships directly into the model architecture, PRIMS bridges analytical fluid mechanics and deep learning, enabling interpretable, data-efficient, and resilient fluid classification. Evaluations on a five-fluid benchmark under dynamic flow, pressure, and temperature conditions show that PRIMS achieves 98.92% average F1-score with only 0.46 million parameters, a 14 times reduction compared to state-of-the-art Transformer-based methods. PRIMS also consistently outperforms prior SOTA models under out-of-distribution shifts to unseen temperature ranges and unseen flow-rate ranges, indicating strong robustness to operating conditions not observed during training. These findings suggest that designing architectures that explicitly mirror governing physical relationships can make them learn transferable, environment-independent representations, improving real-world reliability for microfluidic sensing.
[AI-93] Neptuna: A Comprehensive Machine Learning Framework for Benchmarking Complex Multiphase Flows
链接: https://arxiv.org/abs/2607.22280
作者: Harish Ramachandran,Björn Kimpel,Thomas Paula,Josef Winter,Steffen Schmidt,Nikolaus Adams
类目: Fluid Dynamics (physics.flu-dyn); Artificial Intelligence (cs.AI)
备注:
Abstract:Compressible multiphase flows involving shocks and material interfaces arise in applications such as bubble collapse and droplet breakup, where strong nonlinear interactions produce complex interface deformation, mixing, and multiscale dynamics. Developing reliable machine learning surrogates for these flows remains challenging due to the simultaneous presence of compressibility, sharp discontinuities, and multiphase effects. In this work, we introduce the first large-scale benchmark specifically designed for shock-driven compressible multiphase flows, comprising 2.4 TB of high-fidelity 2D and 3D datasets \footnoteDataset repo: this https URL. Dataset sample videos, this http URL, inference rollout plots from autoregressive rollout of the trained baselines are provided in the supplementary\this http URL featuring shock-induced bubble collapse and droplet breakup. We evaluate diverse surrogate model families on our benchmarking framework: Neptuna \footnoteBenchmarking repo: this https URL, including convolutional, spectral, transformer-based, and pre-trained PDE foundation models. Beyond standard MSE training, we investigate composite losses combining MSE with Sobolev, interface-aware, and structure-aware terms, together with adaptive loss balancing using SoftAdapt and GradNorm. Evaluation includes pointwise, spectral, feature-focused, structural, and physics-informed metrics. Results show that no single model performs best across all datasets and metrics, while composite losses significantly improve interface preservation and spectral fidelity. Among adaptive weighting strategies, SoftAdapt provides the most consistent improvements with almost no overhead compared to MSE-only training.
[AI-94] Explicit Iteration Complexity of Exact Data-Driven Inverse Optimization for Integer Linear Programs
链接: https://arxiv.org/abs/2607.22263
作者: Akira Kitaoka
类目: Optimization and Control (math.OC); Artificial Intelligence (cs.AI); Machine Learning (cs.LG); Machine Learning (stat.ML)
备注: 34 page. This paper was splited from arXiv:2405.14273v7
Abstract:A data-driven inverse optimization problem (DDIOP) is the problem of estimating the objective-function parameters (weights) that explain observed optimal-solution data, and it arises in many applications, including integer linear programming (ILP). It is known that, by applying gradient-based optimization methods to the suboptimality loss, the inverse optimization of ILPs can be solved exactly within finitely many oracle iterations, and that the required number of iterations is bounded as T=O(1/\gamma(\ell_\mathrmsub)^2) in terms of a problem-dependent geometric constant \gamma(\ell_\mathrmsub) . However, no means of bounding \gamma(\ell_\mathrmsub) from below as a function of the problem size has been available, and hence the number of iterations could not be given as an explicit function of the problem size. We therefore give, when the forward problem is an integer linear program (ILP), the number of iterations sufficient for projected subgradient descent applied to the suboptimality loss to achieve exact consistency with the observed data, as a fully explicit function of the number of samples, the dimension of the features, the ranges of the features, and the structure of the constraint coefficient matrix, up to polynomial factors in the basic constants (the diameter of the weight set, the step-size parameter, and the Lipschitz constant of the suboptimality loss).
[AI-95] Multiplicity of Stable Attractors in Disordered Neural Models
链接: https://arxiv.org/abs/2607.22047
作者: Raffaele Marino,Roberto Livi,Antonio Politi
类目: Disordered Systems and Neural Networks (cond-mat.dis-nn); Statistical Mechanics (cond-mat.stat-mech); Artificial Intelligence (cs.AI); Chaotic Dynamics (nlin.CD)
备注: 7 pages
Abstract:We show how large-deviation statistics allows one to obtain reliable estimates of the multiplicity of stable fixed-points in a model of neural ordinary differential equations previously employed in computational tasks. The result is obtained by developing a suitable perturbative method in the amplitude of the disorder. It turns out that for not-too-large coupling strengths there are no qualitative differences between the symmetric case, when the dynamics is a purely gradient evolution, and the asymmetric case, when limit cycles and chaos can, in principle, arise. The selection of this specific model is dictated by pedagogical reasons, but we are confident that the approach can be extended to other many-degree-of-freedom dynamical models characterized by different classes of random coupling matrices.
[AI-96] Interventional Score Geometry for Causal Inference
链接: https://arxiv.org/abs/2607.21914
作者: Mojtaba Eslami
类目: Methodology (stat.ME); Artificial Intelligence (cs.AI); Econometrics (econ.EM)
备注:
Abstract:Let p(x) be the joint density of variables X , and let \psi(x)=\nabla_x\log p(x) be its score field. Geometry constructed from p and \psi alone cannot identify causal direction: structural models with the same observational distribution have the same score geometry. I develop an interventional analogue. A hard intervention \operatornamedo(X_k=\xi) does not merely reweight the joint law; it restricts the distribution to the submanifold x_k=\xi . Its score should therefore be defined on the remaining d-1 free coordinates. I define causal influence X_k\rightsquigarrow X_j as variation of the interventional marginal distribution of X_j with \xi , and show that the corresponding derivative of the marginal interventional score gives a local sufficient condition for influence. Projecting the observational score onto admissible intervention directions does not generally recover causal response: two models may share the same observational score and admissible set yet respond differently. I therefore introduce an interventional response field supplied by structural information. A causal metric is defined as the Fisher information metric on a family of interventions with a common target, avoiding ill-posed comparisons across targets. The framework yields a geometric dictionary for randomized trials, instrumental variables, and conditional-independence designs, clarifying what each does and does not identify. A bivariate Gaussian example gives two models with the same observational score but different interventional score derivatives. The framework organizes relations among designs, interventions, and score fields, but adds no identification beyond the underlying assumptions. In Pearl’s Ladder of Causation, observational score geometry belongs to association, intervention-indexed score fields to intervention, and unit-level counterfactual geometry is left for future work.
[AI-97] Graph-Theoretic Neural Network Frag mentation with Covariant Direct Molecular Force Learning: Enabling Coupled-Cluster Accuracy AIMD for Fluxional Systems
链接: https://arxiv.org/abs/2607.21779
作者: Xiao Zhu,Srinivasan S. Iyengar
类目: Chemical Physics (physics.chem-ph); Artificial Intelligence (cs.AI); Computational Physics (physics.comp-ph)
备注:
Abstract:Accurate ab initio molecular dynamics (AIMD) simulations of complex, fluxional chemical systems are severely limited by the high computational scaling of correlated electronic structure methods. To overcome this bottleneck, we present a robust, graph-theoretic molecular fragmentation framework integrated with machine learning to directly model post-Hartree-Fock nuclear forces at coupled cluster accuracy. Bypassing the limitations of automatic differentiation on learned energy surfaces that may struggle with link-atom Jacobians, our approach directly predicts nuclear force vectors. By projecting these vectors onto fragment-fixed principal axes of inertia, we establish co-variant descriptors that naturally preserve rotational, translational, and permutational invariance. The methodology achieves exceptional high parameter efficiency through a vector-valued training protocol that reduces trainable parameters by over an order of magnitude, while an unsupervised mini-batch k-means space tessellation algorithm constructs highly representative training databases using only 10% to 20% of reference configurations. We rigorously validated this framework on the highly fluxional solvated Zundel cation H_13O_6^+ ). Our fully machine-learning-predicted AIMD trajectories successfully reproduced complex dynamical signatures and key structural characteristics, including radial distribution functions and the velocity autocorrelation power spectrum. Ultimately, this scalable, systematically improvable framework bridges the gap between high-level correlated wavefunction theories and long-timescale reactive sampling, laying the foundation for advanced, LLM-inspired transfer learning in modern chemical dynamics simulations.
[AI-98] Deep Sigma Point Processes for RCS Modeling in Spaceborne SAR Imagery
链接: https://arxiv.org/abs/2607.21745
作者: Khalid El-Darymli,Christoph H. Gierull,Katerina Biron,Weimin Huang
类目: ignal Processing (eess.SP); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注:
Abstract:Radar cross-section (RCS) modeling is foundational to advancing the utility and sensitivity of spaceborne radar systems. This study introduces a deep sigma-point process (DSPP) model for predicting RCS in synthetic aperture radar (SAR) imagery using a RADARSAT-2 dataset containing 208,191 verified ships. The DSPP model not only strives for predictive accuracy but also characterizes the uncertainty inherent in the intricate relationships among radar signals, ship parameters, and environmental conditions. Unlike traditional approaches that rely on deterministic equations with static parameters, the DSPP uses a hierarchical Gaussian process framework with Bayesian inference to capture variability and uncertainty in RCS predictions. By generating predictive distributions rather than single estimates, the model accounts for the complex dynamics governing radar returns. Using a Matern kernel with automatic relevance determination, the DSPP identifies and ranks critical features across radar, operational, and environmental domains, thereby supporting transparency and interpretability. Performance evaluations demonstrate the model’s superiority over linear regression baselines, with a 20.83 percent reduction in root mean squared error, a 25.89 percent increase in R-squared, and a 44.4 percent reduction in both the residual interquartile range and median absolute deviation on the test data. By providing calibrated uncertainty bounds, the DSPP enhances prediction reliability and supports robust decision-making. This work represents a shift toward probabilistic models that incorporate the inherent uncertainty of complex phenomena. By transitioning from fixed equations to distributions over outcomes, the DSPP fosters a deeper understanding of RCS behavior and enables systems to operate effectively in dynamic environments.
[AI-99] Generative and multimodal AI for materials prediction and design: Progress challenges and perspectives
链接: https://arxiv.org/abs/2607.21660
作者: Xianyuan Liu,Charles Anjah,Benjamin E. Jolly,Jonathon F. S. Markanday,Joshua Berry,Haolin Wang,Nicola A. Morley,Robert D. J. Oliver,Alexandra J. Ramadan,Delvin Ce Zhang,Katerina A. Christofidou,Haiping Lu
类目: Materials Science (cond-mat.mtrl-sci); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注:
Abstract:Artificial intelligence (AI) is accelerating materials prediction and design by enabling efficient exploration of chemical and structural spaces, with particular promise for novel materials discovery. However, novelty in materials discovery encompasses chemical plausibility, structural distinctiveness, property relevance and experimental realisability, making AI-driven novelty claims difficult to substantiate. We introduce a materials property hierarchy, from intrinsic, composition-determined properties to extrinsic, processing-dependent performance, to clarify deployment constraints and distinguish structural, physical and deployment novelty. This framework motivates an evidence-based view of multimodal materials data spanning chemical composition, microstructure, processing, and testing and characterisation, showing that current evidence remains concentrated in composition and idealised structure while heterogeneous, under-represented and weakly integrated modalities limit support for physical and deployment novelty. It also highlights the limitations of benchmarks based mainly on computational labels and proxy novelty criteria. Community-wide standards for data collection, modality alignment and evidence synthesis are needed to support multimodal data construction, process-aware multimodal modelling, feasibility-first generative modelling and deployment-aware benchmarking, so that generative and multimodal AI can design experimentally realisable materials with defensible scientific and practical novelty.
[AI-100] Do emulated quantum circuits change what CNNs look at? Performance and explainability comparison in medical image classification
链接: https://arxiv.org/abs/2607.21186
作者: Guillermo Rubiños Rodríguez,Martín Ottavianelli,Mateo Alonso,Gonzalo Blázquez Gil,Boris-Stephan Rauchmann,Pablo Díez-Valle,Sergio Altares-López
类目: Quantum Physics (quant-ph); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: 17 pages, 6 figures, 2 tables
Abstract:Numerous studies have analyzed the use of hybrid quantum-classical convolutional neural networks as a promising alternative to classical deep learning. However, network components on quantum hardware impose fundamental limitations, while the scalability of quantum circuits leads to trainability issues. In this work, we investigate whether small, classically-emulated quantum circuit components can play a meaningful role within complex models, offering an alternative to purely classical convolutional architectures. To this end, we present a systematic study of the effectiveness of a Hybrid Quantum-inspired Convolutional Neural Network (HQiCNN) compared with a parameter-matched classical Convolutional Neural Network (CNN) that differs only in an intermediate dense neural layer. Both models are evaluated on two real-world medical datasets while systematically varying the different hyperparameters, ensuring a fair model comparison that is both dataset and hyperparameter independent. The results show that no architecture consistently dominates the other: the HQiCNN achieves its largest gains in intermediate-data regimes, whereas the CNN reaches the highest accuracies for the largest training sets in both datasets. Furthermore, removing entanglement produces comparable performance while enabling substantially better scalability of quantum simulations, and richer observable sets become beneficial only when sufficient training data are available. Finally, we propose two SHAP-based explainability tools for comparing the predictions between both models, |SHAP| IoU and EMD_pos metric, to demonstrate that both architectures consistently attend to anatomically plausible regions. Thus, we provide a comprehensive benchmark showing that, under certain conditions, hybrid quantum-inspired models are an alternative that can offer benefits in practical tasks such as medical image classification.
机器学习
[LG-0] Dysphagia Risk Stratification in Head and Neck Cancer via Two-Stage PRO-Clinical Stacking
链接: https://arxiv.org/abs/2607.22514
作者: Siyuan Zhao,Eric Ababio Anyimadu,Zachary G. Brumm,Yue Ma,Clifton David Fuller,Xinhua Zhang,G. Elisabeta Marai,Guadalupe Canahuate
类目: Machine Learning (cs.LG)
*备注:
Abstract:Dysphagia is a debilitating late effect of head and neck cancer (HNC) treatment, yet timely identification of at-risk patients remains challenging in survivorship care. Definitive assessment relies on videofluoroscopic imaging, as captured by the Dynamic Imaging Grade of Swallowing Toxicity (CTCAE-DIGEST), which, while validated, requires specialized equipment, trained personnel, and significant patient burden, limiting its routine use in surveillance. Patient-reported outcomes (PROs), by contrast, are low-cost, scalable, and easily collected at any clinical encounter, making them an attractive alternative signal for identifying patients who may warrant further evaluation. However, a clear clinical framework for translating PRO responses into actionable interventions is still evolving. In particular, uncertainty remains regarding when a patient’s self-reported symptom burden should prompt escalation of care. This study addresses this gap by formulating a single-visit PRO-clinical prediction framework and introducing a clinically interpretable two-stage stacking model to predict swallowing impairment risk using PRO responses and structured clinical variables, without requiring videofluoroscopic imaging. The proposed framework quantifies the independent contributions of patient-reported symptoms and clinical factors within a unified and interpretable risk assessment model. Our findings demonstrate that individual MDADI responses contain predictive information beyond that captured by composite or global summary scores, while interpretability analyses reveal symptom patterns and clinical risk factors associated with swallowing impairment. Together, these results support the use of structured PRO-clinical integration as a practical, imaging-free approach for dysphagia risk stratification in HNC survivorship. Subjects: Machine Learning (cs.LG) Cite as: arXiv:2607.22514 [cs.LG] (or arXiv:2607.22514v1 [cs.LG] for this version) https://doi.org/10.48550/arXiv.2607.22514 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[LG-1] Interpretable EEG biomarkers with bag-of-waves: Spatial and temporal waveform dictionaries for low-data regimes
链接: https://arxiv.org/abs/2607.22508
作者: Athanasios Papastathopoulos-Katsaros,Steven T. Lee,Lin Yao,Ajay Thomas,Junseok Park,Matthew J. McGinley,Zhandong Liu
类目: Machine Learning (cs.LG); Signal Processing (eess.SP)
*备注: In Review
Abstract:Electroencephalography (EEG) is widely used to diagnose neurological conditions, but its analysis usually relies on either predefined spectral features or deep neural networks. Predefined features carry a strong bias, since they fix in advance what counts as informative, while deep neural networks and foundation models are hard to interpret and need large amounts of data and compute. We present bag-of-waves, an interpretable framework that learns a small dictionary of recurring EEG waveform templates, called atoms, using shift-invariant k-means without labels. The continuous EEG is then turned into a sequence of atom tokens, whose counts feed a simple downstream classifier or clustering step. We extend this representation in two ways: we add atom-to-atom transitions, which we call n- grams, to capture temporal structure, and we move from single-channel atoms to regional and cross-channel spatial atoms for the multichannel case. We test the method on three complementary datasets, each probing a different aspect: single-channel mouse genotype clustering with only sixteen animals (the low-data and temporal case), resting-state dementia classification (the spatial case), and the TUEV benchmark, a six-way classification of clinical EEG events (a high-data comparison against strong deep and foundation baselines). Across all three datasets, bag-of-waves achieves performance competitive with state-of-the-art deep and foundation models. Yet, it operates with a fraction of the parameter count and provides full interpretability: because every atom corresponds to an inspectable waveform, the method explicitly recovers known clinical morphologies that a neurophysiologist can directly validate. Its main advantage is that it works in the low-data regime where heavier models are a poor fit.
[LG-2] Susceptible Reservoir Architectures for Regime-Conditional Volatility Forecasting
链接: https://arxiv.org/abs/2607.22491
作者: Aliaksei Kaliutau
类目: Machine Learning (cs.LG)
*备注:
Abstract:Volatility forecasting is dominated by persistence and measurement noise, leaving limited residual structure for nonlinear models to exploit. We introduce Susceptible Architectures (SUSA), a reservoir-design principle for volatility forecasting, and its two concrete implementations, based on complex-valued open-chain and periodic reservoirs and regime-conditioned experts to interpret reservoir features across calm, onset, recovery, and persistent-stress states. We also implement open-system q -qubit counterparts in Qiskit while retaining a common AR-Ridge anchor and a bounded residual correction trained under QLIKE. We evaluate models on 16 U.S. equity and exchange-traded-fund series using three disjoint chronological training, validation, and test folds, a 12-observation input window, and a five-observation forecast horizon. The proposed models perform competitively with GARCH, achieving statistically significant QLIKE improvements for specific assets (IWM, XLP). Also models’ forecasts complement HARQ-style predictions: a stacked ensemble improves mean QLIKE by 0.0116 over its strongest constituent and wins in 75% of test scenarios.
[LG-3] Singular value soft-thresholding via the polar decomposition
链接: https://arxiv.org/abs/2607.22484
作者: Stephen Becker
类目: Numerical Analysis (math.NA); Machine Learning (cs.LG); Optimization and Control (math.OC)
*备注: 3 pages, 2 figures, technical report
Abstract:Singular value soft-thresholding can be computed via a reduction to the matrix polar decomposition, which allows one to exploit GPU-friendly algorithms for computing the polar decomposition. Empirically, there is a significant speed-up on GPUs compared to the standard approach using the SVD. We leave the investigation of robustness to future work, but note that due to the discontinuous nature of the sign function, the reduction to the polar decomposition is likely only suitable for low-accuracy applications.
[LG-4] Beyond Negative-Ridge Endpoints: Mixed-Sign Spectral Regularization via Negative-Shifted Gradient Descent
链接: https://arxiv.org/abs/2607.22474
作者: Peng Zhao
类目: Machine Learning (cs.LG); Statistics Theory (math.ST); Machine Learning (stat.ML)
*备注:
Abstract:In overparameterized linear regression, many weak spectral directions act like a ridge penalty on the signal-bearing spectrum; negative ridge is the natural correction, pushing filters above one. The stable negative-ridge endpoint, however, is structurally limited: its pole must stay below the smallest nonzero empirical eigenvalue, and it anti-shrinks smaller eigenvalues more than larger ones. Early-stopped negative-shifted gradient descent escapes this constraint. Its filter is smooth at the would-be pole and mixed-sign-capable: above-ridgeless directions form a leading prefix, with lower directions shrunk or exposure-controlled while stopping sets the crossover. In a Gaussian spike-plus-flat model we discover a Marchenko-Pastur barrier: the shift that cancels the implicit penalty lies a bulk width above the smallest empirical eigenvalue, and the stopped path improves on every admissible endpoint by a polynomial factor in risk under explicit conditions. Our main theorem permits a general high-effective-rank tail: its trace sets the implicit floor, its squared spectrum controls exposure, and the floor-critical path recovers all head scales at once, beyond positive shrinkage and, once scales separate, every uniform rescaling of ridgeless. Handling the noncontractive shifted dynamics is the central technical challenge; localized Duhamel integrals control them. A finite-grid hold-out inequality transfers the separations to the validation-selected algorithm.
[LG-5] Complexity Bounds and Approaches to Learning Projected Gradient Descent Solver Iterates
链接: https://arxiv.org/abs/2607.22467
作者: Anjian Li,Ryne Beeson
类目: Machine Learning (cs.LG)
*备注:
Abstract:Data scarcity poses a fundamental challenge in training generative models to produce initial guesses for parametric optimization problems that are otherwise numerically expensive to solve. We therefore study a k -neighborhood data collection strategy that augments datasets of converged solutions with intermediate solver iterates, increasing the amount of training data without additional solver runs. To understand the benefits of this approach, we derive a generalization bound based on Rademacher complexity that reveals the role of the k -neighborhoods and related parameters. To achieve this result, we focus on one-sided box-constrained quadratic programs solved by projected gradient descent. We illustrate the behavior of this solver on two examples. The approach proposed in this paper enables a more capable DDDAS paradigm by improving the efficiency of the data-model-optimization loop. We finish by discussing two views of learning solver-iterate data and connect our analysis with GLENS, a new data-efficient global search method.
[LG-6] On the Identifiability of Controlled World Models
链接: https://arxiv.org/abs/2607.22430
作者: Xiangteng Zhang,Yang Guan,Bo Zhang,Ya-Qin Zhang,Shengbo Eben Li
类目: Machine Learning (cs.LG)
*备注:
Abstract:Learning world models that infer environment dynamics from high-dimensional observations and predict outcomes under candidate actions is central to planning and control. Joint-Embedding Predictive Architectures (JEPAs) provide a compelling framework for learning such models in representation space. Recent action-conditioned extensions perform promisingly in visual control and latent-space planning, but leave a fundamental question unresolved: when does controlled latent prediction identify both the underlying state and the controlled dynamics? This is challenging under nonlinear observations and behavior policies with limited conditional action variation, where state-dependent evolution and action effects can be statistically confounded. We establish a joint identifiability theory for controlled world models with Gaussian latent states under state-dependent Gaussian behavior policies. We identify two policy-dependent conditions: spectral separation of the predictable signal governs representation identifiability, while non-degenerate conditional action variation governs transition identifiability. We prove that when both conditions hold, every global minimizer of the JEPA objective identifies the latent state and controlled transition up to an orthogonal transformation. We further derive quantitative bounds on representation and transition identifiability under approximate optimization. Finally, we construct predictor perturbations along weakly excited action directions whose counterfactual-to-on-policy error ratio is the inverse transition-identifiability margin, revealing the cost of limited action coverage. Experiments across nonlinear observation maps and behavior policies corroborate the theory and demonstrate implications for transition identifiability, counterfactual prediction, and goal-conditioned latent planning.
[LG-7] LunarFM: A Shared Multimodal Representation of the Moons Surface
链接: https://arxiv.org/abs/2607.22408
作者: Marc Girona-Mata,Jakob Gawlikowski,Sumit Goski,Gautier Bardi de Fourtou,Valentin T. Bickel,Ben Moseley,Abigail Calzada-Diaz,Sylvester Kaczmarek,Raúl Ramos-Pollán
类目: Machine Learning (cs.LG)
*备注: 19 pages, 12 figures
Abstract:The renewed global focus on lunar exploration, driven by the prospect of in-situ resource utilization and a sustained human presence on the Moon, has created growing demand for accurate, large-scale characterization of the lunar surface. Although vast quantities of orbital remote-sensing data have been collected, scientific analysis and resource mapping remain fragmented by heterogeneous multiinstrument observations, sparse labels, and bespoke task-specific modelling workflows. Here we introduce LunarFM, a multimodal foundation model that learns a general representation of the lunar surface from diverse orbital measurements. LunarFM assimilates observations from six instruments across three lunar missions, mapping 18 input channels to a shared embedding space. We demonstrate that this embedding space supports a diverse range of downstream applications, including similarity search, few-shot resource mapping, mineral abundance regression, and geological unit classification, enabling efficient scientific investigation and resource-oriented analysis. We provide a machine-learning-ready dataset of co-registered multimodal observations spanning latitudes from 70°S to 70°N, a pretrained multimodal masked autoencoder, and a companion embedding dataset providing a joint 768-dimensional representation of lunar surface properties. All code and data are available at this https URL
[LG-8] Local-Global Geometric Insights for Graph Neural Networks via Entropic Curvature
链接: https://arxiv.org/abs/2607.22381
作者: Rachid Caich,Yassine Abbahaddou
类目: Machine Learning (cs.LG); Social and Information Networks (cs.SI); Machine Learning (stat.ML)
*备注:
Abstract:Curvature notions on graphs, particularly Ollivier-Ricci and Forman, have emerged as powerful tools for addressing fundamental issues in Graph Neural Networks (GNNs) such as oversmoothing and oversquashing, but rely almost exclusively on local edge-level comparisons and therefore fail to certify how information actually propagates over long distances. We introduce Entropic Curvature, a global, transport-based curvature obtained by extending the Lott-Sturm-Villani framework to graphs through the displacement convexity of entropy along Wasserstein geodesics. We define a tractable Weak Entropic Curvature proxy that lower-bounds the global entropic curvature, and from it derive (i) a Poincare-type inequality controlling oversmoothing, (ii) a transport-entropy generalization bound, and (iii) an expansion paradox proving that sparsity, strong spectral expansion, and positive entropic curvature cannot coexist in large graphs, unifying oversmoothing and oversquashing as opposite ends of a single curvature spectrum. We translate the theory into three practical mechanisms, the E-Gate aggregator, the ENT structural encoding, and Midpoint-Completion Rewiring (MCR), and benchmark them against SDRF, FoSR, BORF, LCP, and Graph Ricci Flow on six node-classification benchmarks, and graph-classification.
[LG-9] Integrated Order Dispatching and Routing for Last-Mile Pickup via Deep Reinforcement Learning
链接: https://arxiv.org/abs/2607.22356
作者: Yida Xu,Zhaofang Mao,Yuheng Miao,Jiaxin Zhang,Yiting Sun
类目: Machine Learning (cs.LG)
*备注:
Abstract:In recent years, the growing complexity of last-mile pickup operations has increased the need for fast and accurate decision-making on logistics platforms. This challenge is fundamentally driven by two key and tightly coupled decision-making processes: order dispatching and routing. Solving them separately overlooks their interdependence, while fully end-to-end learning can be unstable and costly on large, variable-scale instances due to sparse rewards. To solve this problem, we propose an integrated optimization framework which couples a learned routing oracle with real-time dispatching heuristics. For the routing subproblem, we develop a Dynamic-Residual Graph Attention Network encoder with a Look-Ahead Courier-Personalized decoder. For the dispatching subproblem, we develop a routing-oracle-guided dispatching heuristic with local search, where the oracle provides near-optimal solutions to select candidate couriers while retaining real-time scalability. Extensive experiments on real-world datasets from Cainiao Logistics are used to test the performance of our approach, including an offline evaluation and an online rolling-horizon simulation. The experimental results show that our approach outperforms other benchmarks regarding solution quality and solving time, indicating it can effectively support logistics companies in solving real-time and large-scale last-mile pickup problems.
[LG-10] IQ-JEPA: A Joint-Embedding Predictive Architecture with a Hermitian Vision Transformer for Sound Speed and Attenuation Estimation from Ultrasound IQ Data
链接: https://arxiv.org/abs/2607.22351
作者: Masashi Sode,Gianmarco Pinton
类目: Machine Learning (cs.LG); Medical Physics (physics.med-ph)
*备注:
Abstract:The speed of sound in tissue is a prerequisite for well-focused imaging and has diagnostic value, but recovering it from raw pulse-echo channel data is fundamentally a nonlinear inverse problem. Learned solvers are fast yet label hungry. Simulated sound-speed labels are expensive, while abundant real channel data is unlabeled. We propose IQ-JEPA to exploit both data types. An encoder is pretrained without labels to predict the latent representation of masked in-phase and quadrature (IQ) regions from visible context, then fine-tuned on simulated maps. Sound speed appears in the IQ signal as a phase difference, invariant to the constant phase offset. The encoder is a Hermitian vision transformer that operates on the complex signal directly. Its attention is equivariant to that phase and its conjugate-product feed-forward is invariant to it, so the encoder reads a quantity analogous to the one classical coherence methods use. On 79,293 Fullwave 2.5 simulations at 2.5 MHz, pretraining on the 63,435 unlabeled acquisitions reaches 15.60 m/s at 10,000 labels. This is a roughly threefold gain in label efficiency over supervised training, growing to over fourfold at 1,000 labels. It is about 2.2x below an InversionNet baseline, and 8.71 m/s at full labels. The gain still grows with more unlabeled pretraining data. Our comparisons point to self-supervision as the dominant factor. The same encoder transfers. Its frozen features expose sound speed and attenuation, and cross-distribution pretraining between layered and abdominal phantoms costs little accuracy. We see this as a first step toward a foundation model for quantitative ultrasound.
[LG-11] Beyond Binary Rooftop Mapping: A Four-Class Deep Learning Framework for Green Roof Potential Assessment from Open Swiss Geospatial Data
链接: https://arxiv.org/abs/2607.22342
作者: Htet Yamin Ko Ko
类目: Machine Learning (cs.LG)
*备注:
Abstract:The development of effective urban climate adaptation strategies requires comprehensive spatial information on rooftops and buildings, since such information underpins the assessment of ecosystem services provided by green infrastructure, particularly for urban heat island (UHI) mitigation. Although green roofs are widely acknowledged as a promising measure for improving urban thermal comfort, most existing research maps either current green rooftops or rooftops with greening potential, but not both. This study presents a modified deep convolutional neural network rooftop classification framework based on Roofpedia, developed by the Urban Analytics Lab at the National University of Singapore. The proposed model combines high resolution aerial imagery with rooftop slope information derived from a digital surface model and relies entirely on publicly available Swisstopo datasets: SWISSIMAGE orthophotos, swissSURFACE3D elevation data, and swissTLM3D building footprints. Applied to Bern, Switzerland, the model labels rooftops into four categories: existing green roofs, rooftops suitable for green roof installation, rooftops with solar panels, and flat rooftops unsuitable for greening. The framework identifies realistic opportunities for green roof expansion and supplies urban planners with evidence-based information for green infrastructure deployment in Bern and other Swiss cities. Because it is fully open source, the framework is transferable to cities worldwide.
[LG-12] Evolution-Aware MSA Reasoning for Subsampling via Factor Graphs
链接: https://arxiv.org/abs/2607.22314
作者: Zhangzhi Xiong,Minzhang Li,Haotian Yu,Sixian Shen,Kexin Zhang,Mingrui Li,Jie Zheng,Kewei Tu,Jingyi Yu
类目: Machine Learning (cs.LG)
*备注:
Abstract:Multiple Sequence Alignments (MSAs) provide protein language models with explicit evolutionary context, but their large depth makes subsampling unavoidable under limited token budgets. Existing strategies, including random selection, identity-based filtering, and diversity-driven sampling, are effective heuristics, yet provide limited control over the evolutionary signals retained in the subset. In this work, we recast MSA subsampling as an explicit optimization problem, where key evolutionary measures, including query identity and diversity, are treated as controllable objectives. Building on this view, we introduce AP-REASONER, an Affinity-Propagation-based factor-graph approach. With evolution-aware unary factors, exemplar-consistency factors, and two control knobs, AP-REASONER performs factor-graph reasoning through message passing to infer a fixed-budget MSA subset. Experiments on long-range contact prediction and conformational ensemble prediction show that AP-REASONER outperforms baseline subsamplers on structure-sensitive downstream tasks and enables controllable recovery of alternative protein conformations. These results highlight the value of modeling MSA subsampling as a controllable optimization problem, where factor-graph reasoning offers an effective alternative to heuristic selection.
[LG-13] Synthetic Speech Real Signal: Paralinguistic Preservation and Cross-Lingual Augmentation via Voice Cloning INTERSPEECH2026
链接: https://arxiv.org/abs/2607.22304
作者: Roseline Polle,Owen Parsons,George Fairs,Luis Miguel San Martin Fernandez,Cole Looney,Xiaoliang Wu,Alexandra Livia Georgescu,Stefano Goria
类目: Machine Learning (cs.LG); Sound (cs.SD)
*备注: 5 pages, 3 figures. Accepted at Interspeech 2026
Abstract:Synthetic data augmentation in speech is common practice for linguistic tasks like ASR, but has seen far less work for paralinguistic ones, especially clinical tasks where labelled data is expensive and some patient groups are underrepresented. Voice cloning is one such augmentation approach, but is typically evaluated on speech intelligibility (WER) or speaker similarity (SS) rather than on downstream performance, and it remains unclear whether these preserve the paralinguistic signal such tasks depend on. We benchmark eight voice cloning models on five paralinguistic tasks across public and clinical datasets, showing most preserve signal with modest degradation. We then clone English clinical speech into Japanese and find that training on cloned data outperforms raw cross-lingual transfer for depression and anxiety detection on real Japanese speech, suggesting voice cloning is a promising direction for augmenting clinical speech data in low-resource languages.
[LG-14] Efficient Recommendations via Graph Coarsening and Label Propagation
链接: https://arxiv.org/abs/2607.22287
作者: Alessandro Sbandi,Federico Siciliano,Fabrizio Silvestri
类目: Machine Learning (cs.LG)
*备注:
Abstract:Graph-based recommendations are widely adopted in real-world industrial applications. However, graphs in these systems often reach a massive scale, posing notable scalability and efficiency challenges. This requires techniques that can effectively balance predictive quality with computational cost. One promising approach is graph coarsening, an adaptive graph reduction technique that offers a way to systematically construct smaller, yet structurally representative, versions of the original large-scale graphs. In this work, we propose a flexible two-stage diffusion framework that combines graph coarsening with multi-step label propagation in the telecommunications domain. Domain-specific heuristics are applied to first aggregate nodes into meaningful communities, reducing graph size while preserving essential business-relevant relationships. An initial diffusion process done by a Label Propagation Algorithm (LPA) or a Graph Neural Network (GNN) propagates labels across the coarsened graph to produce coarse-grained predictions. Finally, a second LPA within subgraphs generates the final recommendations for individual users. On a real-world telecommunications dataset, when using LPA in both stages, our method achieves up to +24% NDCG@5 over the full-graph LPA baseline. Incorporating a lightweight GNN in the first stage further boosts NDCG@5 by more than 50%, but requires substantial training and inference time. Through extensive experiments and a detailed ablation, we quantify these trade-offs and demonstrate that our coarsening-driven approach delivers an optimal balance between scalability, latency, and recommendation quality. Subjects: Machine Learning (cs.LG) Cite as: arXiv:2607.22287 [cs.LG] (or arXiv:2607.22287v1 [cs.LG] for this version) https://doi.org/10.48550/arXiv.2607.22287 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[LG-15] An Insight on Evaluation Metrics Under the Imbalanced Case of Anomaly Detection
链接: https://arxiv.org/abs/2607.22286
作者: Romain Hermary,Nesryne Mejri,Djamila Aouada
类目: Machine Learning (cs.LG); Machine Learning (stat.ML)
*备注: Published in EUVIP 2026
Abstract:Anomaly detection is inherently characterised by severe class imbalance, making the interpretation of evaluation metrics challenging. Although metrics such as AUROC, AUPR, F1-score, and MCC are widely used, their values convey different meanings depending on the anomaly ratio. In this work, we analyse the behaviour of those four common anomaly detection metrics under varying levels of imbalance. We focus on the study of metric landscapes, visualisations that relate metric values to true positive and true negative rates, providing an intuitive view of metric preferences and stability. Our analysis offers practical guidance for interpreting and comparing anomaly detection results across datasets with different imbalance ratios.
[LG-16] Autoregressive EHR Foundation Models with Multimodal Inputs ALT
链接: https://arxiv.org/abs/2607.22264
作者: Yuxuan Liu,Joshua Placidi,Jinpei Han,Alfred John Balston,Marek Rei,A. Aldo Faisal
类目: Machine Learning (cs.LG)
*备注: 5 pages excluding references and supplements, 2 figures and 2 tables, Proceedings of the Workshop on Structured Data for Health at the 43rd International Conference on Machine Learning, Seoul, South Korea
Abstract:Autoregressive foundation models trained on tokenized electronic health records (EHRs) can support zero-shot clinical prediction, yet most operate on structured event codes alone, and do not incorporate multiple modalities in a principled way. We present a framework for conditioning such models on auxiliary clinical modalities, including ECG waveforms, chest X-ray images, and clinical notes, using modality-specific latent compression and gated cross-attention with temporal alignment. We investigate two key design choices: (1) how to compress long per-modality sequences (e.g., ECG time series) before they enter the multi-modal cross-attention. This feature may be essential to reduce compute overheads and may be beneficial for generalization; (2) how the choice of pretrained encoder for each modality impacts downstream performance. Through controlled ablations on MIMIC-IV, we show that the best latent-compression configurations outperforms both uncompressed cross-attention and mean pooling. Encoder choice has a clear within-modality effect, with stronger pretrained encoders consistently outperforming weaker alternatives. We further show that merely adding auxiliary modalities does not guarantee improvement on ICU mortality prediction over an EHR-only baseline. This implies that careful design of the fusion architecture and an appropriate evaluation in the clinical context are required.
[LG-17] Latent PDE mapping for efficient physics-informed learning across geometries with limited data
链接: https://arxiv.org/abs/2607.22215
作者: Ingvild Askim Adde,Mary M. Maleckar,Gabriel Balaban
类目: Machine Learning (cs.LG); Computational Physics (physics.comp-ph)
*备注:
Abstract:In this study, we introduce latent PDE mapping, a broadly applicable physics-informed learning technique designed to enable efficient geometric generalization with sparse training data. Latent PDE mapping pulls back geometry-specific PDE residuals and boundary conditions to a predefined latent geometry via the deformation gradient, thereby enabling the automated calculation of geometry-consistent shape gradients that are missing in conventional physics-informed machine learning formulations. We demonstrate the utility of latent PDE mapping in solving the anisotropic Aliev-Panfilov PDE of cardiac electrophysiology using both physics-informed neural networks and physics-informed deep operator networks. The Aliev-Panfilov PDE serves as a challenging exemplar: a nonlinear, time-dependent PDE benchmark with sharp gradients that are expensive to capture using traditional numerical solvers. To represent the limited data regime, we train the networks using just fifteen geometric samples drawn from parameterized distributions in two and three spatial dimensions. While modest improvements appear for geometries parameterized by affine and shear deformations, latent PDE mapping demonstrates significant benefits on select geometric families, achieving a factor ~4-6 reduction in mean relative L2 error. Furthermore, our results show that the computational cost of applying latent PDE mapping was modest during network training, and negligible at inference. Taken together, our study highlights how latent PDE mapping facilitates the creation of generalizable physics-informed machine learning models from limited sets of training geometries.
[LG-18] rajectory-Regularized Stochastic Optimal Control via KL Divergence
链接: https://arxiv.org/abs/2607.22201
作者: Mintae Kim,Koushil Sreenath
类目: ystems and Control (eess.SY); Machine Learning (cs.LG)
*备注: 8 pages, 4 figures, 65th IEEE Conference on Decision and Control
Abstract:We introduce trajectory-regularized stochastic optimal control (TRSOC), which augments standard stochastic optimal control (SOC) with a Kullback–Leibler (KL) divergence between controlled and reference trajectory distributions. Using Girsanov’s theorem, the trajectory KL reduces to a quadratic drift mismatch penalty, yielding a modified running cost that preserves the dynamic programming (DP) structure. We derive the corresponding Hamilton–Jacobi–Bellman (HJB) equation and characterize the optimal policy. In the linear-quadratic (LQ) setting, the formulation admits a closed-form solution with an augmented control cost. Experiments show that the regularization parameter induces a trade-off between performance-driven and reference-preserving behavior, including cases with reference dynamics learned from offline data.
[LG-19] From Score Approximation to Distribution Approximation in Score-Based Diffusion Models
链接: https://arxiv.org/abs/2607.22199
作者: Lan V. Truong
类目: Machine Learning (cs.LG); Information Theory (cs.IT); Machine Learning (stat.ML)
*备注: 18 pages
Abstract:Score-based diffusion models have achieved remarkable empirical success in generative modeling, yet their approximation-theoretic foundations remain incomplete. In particular, although classical universal approximation theorems guarantee that neural networks can approximate score functions, it remains unclear whether such approximation guarantees translate into approximation of the probability distributions generated by reverse diffusion processes. In this paper, we establish a rigorous quantitative connection between these two notions. Specifically, we prove that if a neural network approximates the true score function sufficiently accurately, then the probability distribution generated by the corresponding reverse diffusion model is close to the target data distribution in Kullback-Leibler (KL) divergence, up to an irreducible mismatch between the terminal distribution of the forward diffusion process and the prior used to initialize the reverse process. More precisely, we derive an explicit upper bound on the distribution approximation error in terms of the score approximation error, the diffusion noise schedule, and the terminal prior mismatch. Our analysis combines Hornik’s universal approximation theorem, Girsanov’s theorem on path space, and the data processing inequality for relative entropy. Complementary to recent work that studies score approximation under finite-sample statistical settings and structural assumptions on the data distribution, our work develops an approximation-theoretic analysis based on classical neural network approximation theory. The resulting theorem provides a simple and explicit guarantee linking neural network approximation of score functions to approximation of the probability distributions generated by reverse diffusion models.
[LG-20] Unbiased Open World Regularization for Fair Self-Supervised Learning
链接: https://arxiv.org/abs/2607.22149
作者: L{é}o Nicollier(CB, ATT),Marc Pic(ATT),Pablo Mus{é}(CB, IFUMI),Enric Meinhardt-Llopis(CB),Gabriele Facciolo(CB)
类目: Machine Learning (cs.LG)
*备注:
Abstract:Despite recent advances, self-supervised learning (SSL) models and Joint-Embedding Predictive Architectures (JEPAs) remain susceptible to learning spurious biases in the dataset. These techniques rely on regularization, which prevents representation collapse by enforcing a global target distribution such as a multivariate Gaussian or a uniform distribution on the sphere. However, these global constraints are insufficient to prevent bias entanglement, as task-irrelevant features can still segregate the latent space into distinct sub-regions. While recent approaches like Entangling and Disentangling (EnD) and Fair Supervised Contrastive Learning (FSCL) empirically debias the latent space, we show that they act as partial approximations of conditional distribution matching. To enforce this matching explicitly, we propose Unbiased Open World Regularization (UOWReg), an encoder-only framework. We show that this shift from a global to a conditional objective guarantees statistical independence between the learned representations and the targeted attributes, regardless of the chosen target distribution. We empirically validate this framework across both Gaussian and spherical latent spaces, using statistical measures to enforce these target distributions. While conditional matching successfully mitigates bias with both distributions, we demonstrate that enforcing conditional uniformity on the sphere yields a lower linearprobing classification error. Empirically, UOWReg reduces Equalized Odds violations on the CelebA benchmark while maintaining competitive classification accuracy compared to existing encoder-only baselines. Furthermore, we introduce the Synthetic Engraving Task-a novel setting in which a dominant macro-structure masks a fine-grained micro-signature. We show that UOWReg effectively prevents the subpopulation collapse observed in standard SSL, successfully isolating micro-signatures even when heavily entangled with the global structure.
[LG-21] Pretraining EHR Foundation Models with Patient-Aware Sampling ALT ICML2026
链接: https://arxiv.org/abs/2607.22114
作者: Joshua Placidi,Yuxuan Liu,Jinpei Han,Marek Rei,A. Aldo Faisal
类目: Machine Learning (cs.LG)
*备注: Accepted at the Workshop on Structured Data for Health at the 43rd International Conference on Machine Learning (ICML 2026). 7 pages, 4 figures
Abstract:Autoregressive foundation models for electronic health records (EHRs) typically inherit pretraining methods from language modeling, where patient trajectories are concatenated into a single token stream and windows are sampled from that stream. In EHR data, this choice is consequential: windows may mix multiple patients, and patients with longer records contribute more optimization updates, potentially introducing bias. We propose Patient Sampling, a pretraining sequence-construction method that allows us to control how training signal is distributed across patients. We compare this method to the standard approach, which we refer to as Global Stream. We show that stochastic Patient Sampling with controllable weighting improves performance on real-world EHR data. Across downstream clinical tasks on MIMIC-IV v2.2 and v3.1, Patient Sampling improves Macro AUROC and AUPRC over the Global Stream baseline. These results identify training and validation sequence construction as important and underexplored design choices for autoregressive EHR foundation models.
[LG-22] MemNMF: Memory-Augmented NMF on LPC Spectra for Anomalous Sound Detection
链接: https://arxiv.org/abs/2607.22086
作者: Phurich Saengthong,Takahiro Shinozaki
类目: ound (cs.SD); Machine Learning (cs.LG)
*备注:
Abstract:Autoencoder-based anomalous sound detection is attractive for machine condition monitoring because it can be trained using only normal recordings and yields an interpretable anomaly score from reconstruction error. Most prior work uses spectrogram autoencoders, but reconstructing detailed time–frequency patterns is sensitive to noise and transients, and models can reconstruct some anomalous inputs well, weakening normal–anomaly separation. We propose MemNMF, a constrained reconstruction method that operates on the Linear Predictive Coding spectrum, a compact estimate of the spectral envelope. MemNMF initializes a memory module from an NMF dictionary learned on normal LPC spectra and reconstructs each input as an attention-weighted combination of prototypical normal spectral patterns. Experiments on MIMII and DCASE 2020 Task 2 across multiple machine types and operating conditions show that LPC-spectrum inputs improve a standard autoencoder baseline and that MemNMF yields further gains, with especially strong robustness under noisy, non-stationary settings.
[LG-23] DCS: A Unified Conditional Sensitivity Framework for Cross-Modal Copyright Infringement Detection
链接: https://arxiv.org/abs/2607.22035
作者: Xiafeng Man
类目: Machine Learning (cs.LG)
*备注:
Abstract:Currently, most foundation models can reproduce or strongly depend on copyrighted training content, but output similarity alone is insufficient for infringement detection, because similar outputs may also arise from public-domain concepts, common stylistic conventions, or ordinary statistical generalization. In this paper, we develops a unified post-hoc detection framework that treats copyright infringement evidence as a counterfactual conditional distribution shift: a protected target is suspicious when the model’s behavior under aligned conditions would change measurably if that target were included in, or removed from, the training process. We formalize this view through conditional differential privacy and introduce Dual-Branch Conditional Sensitivity (DCS), an operational statistic that measures the observable gap between two locally perturbed model states. Specifically, the proposed DCS framework creates a learning branch and an unlearning branch around the deployed model, connects their displacement to the unavailable counterfactual retraining effect through influence-function analysis, and bounds the observable sensitivity by the counterfactual privacy-budget surrogate, local curvature, training-set scale, and perturbation step size. To distinguish target-specific memorization from generic fine-tuning instability, we further define a calibrated detection statistic that subtracts the sensitivity measured under orthogonal conditions. The DCS framework is instantiated for ridge-regularized linear regression, conditional diffusion models, autoregressive language models, and multimodal models. These instantiations show how the same principle can be evaluated through prediction gaps, image-embedding divergence, token-distribution or entropy shifts, and cross-modal representation changes. Subjects: Machine Learning (cs.LG) Cite as: arXiv:2607.22035 [cs.LG] (or arXiv:2607.22035v1 [cs.LG] for this version) https://doi.org/10.48550/arXiv.2607.22035 Focus to learn more arXiv-issued DOI via DataCite (pending registration) Submission history From: Xiafeng Man [view email] [v1] Fri, 24 Jul 2026 07:00:44 UTC (1,170 KB)
[LG-24] Cross-Domain Off-Policy Evaluation and Learning for Contextual Bandits ICLR2025
链接: https://arxiv.org/abs/2607.22012
作者: Yuta Natsubori,Masataka Ushiku,Yuta Saito
类目: Machine Learning (cs.LG)
*备注: 21 pages, 10 figures, accepted to ICLR 2025
Abstract:Off-Policy Evaluation and Learning (OPE/L) in contextual bandits is rapidly gaining popularity in real systems because new policies can be evaluated and learned securely using only historical logged data. However, existing methods in OPE/L cannot handle many challenging but prevalent scenarios such as few-shot data, deterministic logging policies, and new actions. In many applications, such as personalized medicine, content recommendations, education, and advertising, we need to evaluate and learn new policies in the presence of these challenges. Existing methods cannot evaluate and optimize effectively in these situations due to the notorious variance issue or limited exploration in the logged data. To enable OPE/L even under these unsolved challenges, we propose a new problem setup of Cross-Domain OPE/L, where we have access not only to the logged data from the target domain in which the new policy will be implemented but also to logged datasets collected from other domains. This novel formulation is widely applicable because we can often use historical data not only from the target hospital, country, device, or user segment but also from other hospitals, countries, devices, or segments. We develop a new estimator and policy gradient method to solve OPE/L by leveraging both target and source datasets, resulting in substantially enhanced OPE/L in the previously unsolved situations in our empirical evaluations.
[LG-25] Energy Manifold Natural Gradient Descent: Riemannian Optimization for Neural PDE Solvers
链接: https://arxiv.org/abs/2607.22004
作者: Zhangyong Liang,Huanhuan Gao
类目: Machine Learning (cs.LG)
*备注:
Abstract:Energy natural gradient descent (ENGD) aligns parameter updates with the curvature of an underlying function-space energy, but existing formulations assume an unconstrained Euclidean parameter domain. We introduce \EMNGDfull, a manifold optimization framework for physics-informed and variational neural PDE solvers whose parameters lie on a Riemannian manifold. EMNGD restricts the energy-induced quadratic model to feasible tangent directions and uses retractions to preserve parameter constraints throughout optimization. Under coercivity, we prove that the push-forward of the undamped EMNGD direction is the best feasible approximation to the function-space Newton vector in the energy metric. We establish coordinate invariance, exact reduction to ENGD in Euclidean space, global first-order convergence with Armijo backtracking, and robustness to inexact tangent solves. For quadratic residual energies and generalized Gauss–Newton pullbacks, the Woodbury identity transfers the tangent system to sample space without changing the direction. Nyström approximation provides scalable sample-space solves with controlled direction error and recovers the exact direction after iterative convergence. On the evaluated neural PDE benchmarks, EMNGD achieves higher accuracy and faster convergence than the compared state-of-the-art baselines. Woodbury preserves the EMNGD direction, while scalable-solver diagnostics quantify the accuracy–cost trade-off of preconditioning and residual subsampling.
[LG-26] On the Convergence of Stochastic Low-Rank Adaptation
链接: https://arxiv.org/abs/2607.21975
作者: Ru Wang,Chengchang Liu,John C.S. Lui
类目: Machine Learning (cs.LG); Optimization and Control (math.OC)
*备注:
Abstract:Low-rank adaptation (LoRA) optimizes J(B,A)=\mathcal L(W_\mathrmbase+sBA) over two adapters B \in \mathbbR^m \times r and A \in \mathbbR^r \times n that form a low-rank update to a frozen pretrained weight matrix W_\mathrmbase \in \mathbbR^m \times n . The prior analysis shows LoRA-GD takes \exp\mathcalO(\epsilon^-2)\ oracle calls to find an \epsilon -stationary point such that |\nabla J(B,A)|\leq \epsilon in the deterministic setting. We sharpen the analysis and show that \mathcalO(\epsilon^-4) full-gradient evaluations suffice for the same first-order criterion. We further study stochastic LoRA under unbiased gradient estimates and finite variance. We propose LoRA-NSGDM, which finds an \epsilon -stationary point with \mathcalO(\epsilon^-8) stochastic oracle complexity. Under the additional mean-square smoothness condition, we use variance reduction strategy and propose LoRA-STORM, which improves the stochastic oracle complexity to \mathcalO(\epsilon^-6) .
[LG-27] Multi-Agent Debate and Visual Information Extraction for SeePhys Pro: A 1st-Place Technical Report from ICML 2026 AI4Math Track 3 Challenge
链接: https://arxiv.org/abs/2607.21946
作者: Jiseok Kwak,Suhyeon Jo,Taewoo Kim,Yeongmin Kim,Byeonghu Na,Il-chul Moon
类目: Machine Learning (cs.LG)
*备注:
Abstract:This technical report presents our approach to Challenge Track~3: SeePhys Pro at the 3rd AI for Math Workshop, where the task is to answer college-level physics questions whose statement and figure may be given partly or entirely as an image. Visual physics problems become substantially harder for large language models when the decisive information resides in a figure rather than in the text, and this modality gap widens as more of the problem migrates into the image. We address the task with a two-stage framework: a visual information extraction stage that re-expresses figure content as solver-readable text to close the modality gap, and a reasoning stage that orchestrates three heterogeneous solvers through multi-agent debate. Our analysis yields two findings: the gain from orchestration comes from reliable answer selection rather than from additional debate, and the value of a figure aid scales with how much of the problem is locked inside the image. The resulting pipeline improves overall accuracy over a single-agent baseline from 0.643 to 0.802 on the public split, and won 1st place on both the public and the private leaderboard (private overall 0.743).
[LG-28] RIS-Kernel: A Model-Agnostic Architecture for Long-Context LLM Inference via Sparse Attention
链接: https://arxiv.org/abs/2607.21927
作者: Anderson R. Santos
类目: Machine Learning (cs.LG)
*备注: 20 pages, 9 figures, 5 tables
Abstract:Full self-attention in large language models scales as O(N^2), which limits long-context document analysis to 65,536 tokens and requires costly GPU clusters. The Reduced Interaction Sampling (RIS) inference engine addresses this constraint as a model-agnostic architecture. Without modifying weights, RIS reduces self-attention complexity to O(N log N) using sparse stochastic geometry that fits within commodity memory limits. We validate RIS on Qwen2-1.5B-Instruct across two regimes. In controlled evaluations at 32,768 tokens (where native dense attention serves as the upper bound), RIS-Stochastic at 1% density and 70 ensemble seeds achieves 75.00% accuracy, outperforming the native dense baseline (71.88%), while RIS-Stochastic at 5% density and 10 seeds matches it (71.88%). This demonstrates that sparse attention acts as a regularizer: low density (1%) over multiple seeds filters out sequence-level noise, whereas higher density (5%) reintroduces distractor noise. Under the tightest budget, RIS-Structural reaches 68.75% accuracy at 1% density with just 10 seeds, recovering 75% of the contextual gap relative to the zero-context floor (59.38%). At 65,536 tokens, where dense attention triggers out-of-memory faults, RIS yields retrieval gains of up to 14.06 percentage points over the zero-context floor (51.56%), which is confirmed as marginally significant under McNemar’s paired test (p = 0.078 0.10). All evaluations run on commodity, unaccelerated CPU servers (16-128 GB of RAM), demonstrating that long-context LLM inference is feasible on standard academic hardware without GPU acceleration.
[LG-29] MissHyper: Restoring Clinical Synchronicity in Missingness-Guided Hypergraph Forecasting
链接: https://arxiv.org/abs/2607.21922
作者: Mingyi Ma,Qingxiong Tan
类目: Machine Learning (cs.LG)
*备注:
Abstract:Clinical irregular multivariate time series are shaped not only by physiological dynamics but also by the measurement process that determines when and what to observe. In event-centric models, however, co-timestamp structure can be flattened too early: measurements acquired at the same timestamp are embedded as isolated nodes, leaving local patient-state context unavailable until later message-passing layers. We study this pre-propagation representation bottleneck and address it by restoring co-timestamp context before message passing begins. We propose MissHyper, a missingness-guided hypergraph forecasting model with pre-propagation synchronicity restoration. MissHyper augments each event with a local support-density cue, aggregates co-timestamp records to recover patient-state context, and uses a missingness-guided gate to adaptively fuse node-specific evidence with the recovered context. Across PhysioNet 2012, MIMIC-III, and MIMIC-IV, MissHyper achieves consistent gains in multi-step forecasting and outperforms a strong hypergraph baseline. These results suggest that improving event initialization can benefit sparse clinical forecasting without requiring a redesigned downstream propagation architecture. Ablations indicate that snapshot restoration, adaptive fusion, and support-density encoding all contribute, pointing to event initialization as a critical design axis for sparse clinical forecasting.
[LG-30] Remedying Coarsening-Based GNN Training under Heterophily via Adaptive Complementary Enhancement UAI2026
链接: https://arxiv.org/abs/2607.21885
作者: Guoming Li,Jian Yang,Xukun Wang,Zixiao Wang,Shangsong Liang,Yifan Chen
类目: Machine Learning (cs.LG); Signal Processing (eess.SP); Numerical Analysis (math.NA)
*备注: Accepted at the 42nd Conference on Uncertainty in Artificial Intelligence, UAI 2026
Abstract:Coarsening-based training for graph neural networks (GNNs), i.e.\ training on coarsened graphs rather than the original large ones, has become a promising direction for scaling GNNs to massive graphs. However, prior work has been evaluated almost exclusively on \textithomophilic graphs, leaving the more challenging \textitheterophilic settings underexplored. We show, both empirically and theoretically, that existing coarsening-based training methods suffer significant performance degradation on heterophilic graphs due to inevitable loss of graph information during coarsening. To address this, we propose \bf Adaptive \bf Complementary \bf Enhancement, a plug-and-play, model-agnostic strategy that reintegrates the information discarded in coarsening: ACE learns a projector for re-constructing original node features and applies \textitanisotropic structural regularization to embed local heterophily. We further adopt \textithomoscedastic uncertainty weighting to adaptively balance the combined training objective of primary coarsened-graph training loss and full-graph auxiliary loss with augmented node features re-constructed by the heterophily-aware projector. Extensive experiments show that ACE drives consistent gains on heterophilic benchmarks while preserving competitive results on homophilic graphs with minimal computational overhead. Code is available at the GitHub repository: this https URL.
[LG-31] Variance-Reduced Q-Learning over Static and Time-Varying Networks
链接: https://arxiv.org/abs/2607.21876
作者: Sreejeet Maity,Feng Zhu,Aritra Mitra,Robert W. Heath Jr
类目: Machine Learning (cs.LG); Systems and Control (eess.SY)
*备注: Accepted at the 2026 American Control Conference (ACC 2026)
Abstract:We investigate a decentralized reinforcement learning problem involving multiple agents that interact with the same Markov Decision Process (MDP). The agents can exchange information over a network to collectively learn the optimal state-action value function. For this setting, we introduce a novel epoch-based distributed Q -learning algorithm called VRDQ, where within each epoch, agents locally estimate the Bellman optimality operator and diffuse information using a consensus-based protocol. For both static and time-varying networks, we establish high-probability finite-time convergence rates for VRDQ that enjoy linear speedups from collaboration. Crucially, we prove that such speedups in sample-complexity require only \tildeO(1) communication, substantially improving upon the communication costs in prior work.
[LG-32] Scaling Laws for Classical Machine Learning on Tabular Data: A Benchmark Study
链接: https://arxiv.org/abs/2607.21866
作者: Kaihua Ding
类目: Machine Learning (cs.LG); Machine Learning (stat.ML)
*备注:
Abstract:Prior classical-ML learning-curve work fits power laws to tree, linear, and kernel models on tabular data, but at small scale: typically one curve, one team, a handful of cells. We present a distributed classroom-scale replication: 127 students each ran a fixed protocol on 3 assigned datasets, drawn from 18 tabular classification and regression datasets and 6 model families (Boosting, Random Forest, SVM, Linear/Logistic, Ridge, Lasso), yielding 11,536 training runs and 1,648 fitted power-law curves of the form error(N) = a N^(-b) + c. Three findings. (1) Power laws fit: R^2 0.8 on 77.7% of cells, with tree ensembles dominating at full data (Boosting 50% of datasets, RandomForest 33%; linear models underperform on classification). (2) Approximate shared exponents within a model family: for 5 of 6 families, a single family-level exponent predicts each family’s cross-dataset curves nearly as well as per-dataset exponents (R^2 gap 0.011), though AIC favors the unconstrained fit and curve collapse is partial (32-58% of points within +/-0.5 dex). We frame this as approximate predictive compressibility, not dataset-independent universality; Lasso fails outright (negative control) and Ridge is fragile under leave-one-dataset-out. (3) Replicator-implementation variance: with random_state=42 fixed, independent re-implementations of the same protocol still differ by mean CV(b) = 0.144 on the fitted exponent – not seed variance, but the spread induced by unconstrained parts of the protocol (preprocessing, encoding, missing-value handling). We release the aggregated curves, per-cell fits, and a practical data-requirement table for N* to reach target error 0.15.
[LG-33] Searching the Space of Feed-Forward Neural-Network Weight-Update Rules with Fixed Depth Symbolic Regression
链接: https://arxiv.org/abs/2607.21855
作者: Charles Brum,Edward Finkelstein
类目: Machine Learning (cs.LG)
*备注: 15 pages, 1 figure, 8 tables
Abstract:We investigate whether symbolic regression can discover explicit neural network weight-update rules that outperform standard hand-designed optimizers on small symbolic regression benchmarks. Candidate update rules are represented as fixed-depth symbolic expressions over operands derived from common optimizers, including gradient, momentum, adaptive-gradient, and moment-estimate quantities. Across 30 benchmark/neural network combinations, the symbolic regression procedure found an update rule outperforming the best hyperparameter-tuned established optimizer in 25 cases, with an aggregate MSE reduction of 44.47% over the improved cases. The discovered rules do not all share a single common symbolic form, but many combine adaptive normalization, momentum-like quantities, nonlinear transformations, and rational expressions. These results suggest that symbolic regression can serve as a lightweight mechanism for discovering compact optimizer variants, while also highlighting the need for larger-scale validation.
[LG-34] Quantifying Political Partisanship for Cross-Platform Analyses
链接: https://arxiv.org/abs/2607.21842
作者: Fathima Ameen,Christopher G. Healey
类目: ocial and Information Networks (cs.SI); Machine Learning (cs.LG)
*备注:
Abstract:Research on political polarization on social media depends on the ability to reliably measure partisanship in user-generated content. However, existing approaches are typically tailored to platform-specific properties, such as structural affordances or linguistic conventions, which hurts generalizability across platforms. This limitation is increasingly consequential as the social media ecosystem fragments and fringe, alt-tech platforms emerge alongside mainstream ones. We propose a text-based, platform-portable methodology for measuring political partisanship in social media posts, anchored by an external news-credibility signal. Posts are embedded using a transformer-based sentence encoder and clustered into topic groups, which are labeled using the aggregated AllSides media bias scores of cited news outlets. A partisanship axis is then constructed in the embedding space as the difference between centroids of oppositely labeled clusters, and individual posts are scored by projection onto this axis. We apply the method to a corpus of approximately 1.3 million posts collected from Bluesky and Truth Social during the six months preceding the 2024 U.S. presidential election, providing the first cross-platform comparison of partisanship distributions on these two ideologically asymmetric platforms. The resulting partisanship scores correlate significantly with held-out AllSides media bias scores both in-distribution and out-of-distribution on an independent Twitter corpus, and recover within-platform partisan dynamics that platform identity alone cannot explain.
[LG-35] Certified in Theory Broken in Practice: Assumption Gaps in Cryptographic Model Certification
链接: https://arxiv.org/abs/2607.21839
作者: Carter Luck,Olive Franzese-McLaughlin,Elisaweta Masserova,Akira Takahashi,Antigoni Polychroniadou,Nicolas Papernot
类目: Cryptography and Security (cs.CR); Machine Learning (cs.LG)
*备注:
Abstract:Privacy-preserving machine learning auditing protocols allow auditors to assess models for properties such as accuracy or fairness, without revealing their internals or training data. This makes them especially attractive for auditing models deployed in sensitive domains such as healthcare or finance. For these protocols to be meaningful in real-world audit settings, though, their guarantees must reflect how the model will behave once deployed, rather than merely certifying its behavior during an audit. Existing security definitions often miss this mark: most certify model behavior only on a fixed audit dataset, without ensuring that the same guarantees generalize to other datasets drawn from the same distribution. As we show, this gap allows a model provider to attack many cryptographic model certification (CMC) schemes built on secure zero knowledge proofs (ZKP) by carefully engineering training data, resulting in models that exhibit benign behavior during an audit, but pathological behavior in practice. For example, we empirically demonstrate that an attacker can certify that a model achieves over 99% accuracy on an audit dataset, but less than 30% accuracy on fresh samples from the same distribution. To address this gap, we formalize rigorous cryptographic security notions tailored to CMC frameworks, introduce a generic protocol template, and prove that it satisfies these requirements. Our results thus offer both cautionary evidence about existing approaches and constructive guidance for designing secure, privacy-preserving ML auditing protocols. Subjects: Cryptography and Security (cs.CR); Machine Learning (cs.LG) Cite as: arXiv:2607.21839 [cs.CR] (or arXiv:2607.21839v1 [cs.CR] for this version) https://doi.org/10.48550/arXiv.2607.21839 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[LG-36] How Do AI Coding Agents Contribute to Software Development? an Empirical Study of Agent ic Pull Requests
链接: https://arxiv.org/abs/2607.21832
作者: Iren Mazloomzadeh,Mohammad Mehdi Morovati,Foutse Khomh
类目: oftware Engineering (cs.SE); Machine Learning (cs.LG)
*备注:
Abstract:Recent advances in large language models and their rapid adoption across software engineering tasks have made Artificial Intelligence (AI) coding agents an integral component of modern software development workflows. While developers increasingly benefit from these coding agents, their impact on software quality remains insufficiently understood. In particular, how agentic contributions evolve across the software development lifecycle has not been thoroughly investigated. This study aims to characterize agentic pull requests (PR) in comparison to human generated PRs and to examine how their properties change across different stages of the development lifecycle. Using the AIDev dataset, we first analyze how differences in merge rates between agentic and human generated PRs vary over time. We then identify the types of development tasks where AI coding agents are predominantly applied and investigate how these task distributions evolve across development quarters. Finally, we compare a set of key characteristics of agentic and human generated PRs, focusing on their implications for software quality and their temporal dynamics. Overall, our findings provide an empirical and longitudinal perspective on the role of AI coding agents in software development, offering a more nuanced understanding of their benefits and limitations in real-world practices.
[LG-37] A Graph-Based Control Interface for Traffic Signals on Heterogeneous Road Networks
链接: https://arxiv.org/abs/2607.21831
作者: Bertil Braun
类目: Machine Learning (cs.LG); Systems and Control (eess.SY)
*备注: 9 pages, including 3 appendix pages, 5 figures. Code: this https URL
Abstract:We present a traffic-signal control interface in which a shared graph neural network assigns scores to individual traffic movements. Each junction converts these scores into its own variable-sized set of legal signal phases using a deterministic incidence matrix. Directed corridor nodes provide traffic context, while movement nodes represent controlled input-to-output paths through junctions. Typed mean aggregation produces one scalar per movement; phase definitions and signal timing remain outside the learned network. This makes graph size and junction-specific action count independent of the learned parameter shapes. PPO experiments evaluate the interface on unseen synthetic grid geometries, altered signal coverage, and five heterogeneous city graphs. The policies retained performance across unseen geometries within the synthetic grid family, while changes in signal coverage exposed sensitivity to a signal-coverage distribution shift. A single trained city-policy instance executed across all five city graphs, with heterogeneous outcomes. These results provide feasibility evidence rather than a general estimate of transfer to arbitrary road networks.
[LG-38] Data eccentricity asymptotics of Gaussian RBF reproducing kernel Hilbert space and kernel PCA
链接: https://arxiv.org/abs/2607.21823
作者: Sergio A. Alvarez
类目: Machine Learning (cs.LG)
*备注:
Abstract:We show that, up to isotropic scaling, the Gaussian RBF reproducing kernel Hilbert space (RKHS) is asymptotically isometric to Euclidean space in the large bandwidth limit. This strongly suggests that kernel-based constructions reliant on metric properties of the RKHS will yield results for Gaussian RBF kernels that similarly approach those of linear kernels for large bandwidths. The asymptotic behavior of Gaussian CKA can be understood in this light. We further consider kernel PCA, showing that Gaussian RBF eigenvalues, eigenprojections, and principal components all converge to those of classical (linear) PCA as bandwidth \sigma \rightarrow \infty . For a given data representation, both the RKHS feature embeddings and the orthogonal PCA eigenframes of the two kernel types differ asymptotically by a geometric similarity transformation, up to a residual of size O \left (\frac\rho\sigma \right )^2 , where \rho is a measure of geometric eccentricity of the representation, equal to the ratio of maximum to median pairwise distance between data examples. Experiments over a diverse collection of data sets demonstrate that \rho provides a simple and reliable predictor of dataset-specific convergence behavior in the top principal directions.
[LG-39] Bounding the Causal Impact of ML-assisted Decision-Making via Counterfactual Correctness
链接: https://arxiv.org/abs/2607.21806
作者: Jonathan Zhang,Erik Skalnes,Jacob Chen,Michael Oberst
类目: Machine Learning (cs.LG)
*备注: 45 pages. In proceedings of the 42nd Conference on Uncertainty in Artificial Intelligence
Abstract:Predictive machine learning (ML) models are increasingly used to aid human decision-makers across various high-risk domains such as healthcare and criminal justice. There is a growing recognition of the need to evaluate the causal impact of deploying these systems on downstream outcomes, such as patient survival or crime recidivism. Randomized control trials (RCTs) can provide high-quality evidence on the impact of a deployed model, but they run into a challenge: it is often infeasible to run repeated trials when models are updated or retrained to improve predictive performance. In this work, we present a partial-identification approach to using prior RCT data to construct bounds on the causal effect of a new model. The core innovation in our approach is to leverage assumptions relating fine-grained predictive accuracy to downstream outcomes. We do so via two monotonicity assumptions: first, on individual-level `counterfactual correctness’ (all else being equal, a correct prediction leads to non-inferior outcomes); and second, on the relation between subgroup predictive performance and outcomes, interpretable as an assumption regarding trust in model outputs. We demonstrate our method with a simulation study, illustrating how incorporating this information can lead to more informative bounds compared to prior work.
[LG-40] Relaxed activation analysis of dataflow networks - A clock calculus for machine learning and real-time scheduling
链接: https://arxiv.org/abs/2607.21797
作者: William Gaudelier,Albert Cohen,Dumitru Potop Butucaru
类目: Programming Languages (cs.PL); Machine Learning (cs.LG)
*备注:
Abstract:Previous work has shown that the simple dataflow primitives of the Lustre language allow the natural, semantically unambiguous, and compact representation of machine learning (ML) applications, including models featuring complex conditional execution and recurrent state. The Lustre clock calculus is responsible for the static determination of important properties such as liveness (absence of deadlocks) and static memory bounds. Yet existing clock calculi are tailored for embedded control applications. We show they do not cater for the representation of control patterns commonly found in training algorithms, resulting in cumbersome expressions and inefficient compilation. We propose a conservative extension of Lustre’s clock calculus addressing this limitation, thereby facilitating the embedding of ML models in reactive applications.
[LG-41] Smart predict-then-robustly-optimize
链接: https://arxiv.org/abs/2607.21773
作者: Aakil Caunhye,Xuefei Lu,Belen Martin-Barragan
类目: Machine Learning (cs.LG); Optimization and Control (math.OC)
*备注:
Abstract:In this paper, we propose and study a robust variant of the smart predict-then-optimize approach that accounts for prediction shifts due to disturbance in the covariate feature space. While traditional integrated-learning-and-optimization models assume that side information is perfectly revealed, empirical data-driven features are frequently corrupted or noisy at the time of decision-making, leading to fragile operational policies. To bridge this gap, we integrate principles of robust optimization directly into the predictive-prescriptive pipeline via a smart predict-then-robustly optimize loss and establish a computationally tractable convex surrogate, designed to hedge against worst-case feature perturbations. On the theoretical front, we formalize the structural validity of this surrogate by proving its approximation error probability decays exponentially according to a sub-Gaussian concentration profile. Furthermore, we establish that under mild assumptions, the surrogate is Fisher consistent with high probability. We also prove necessary conditions under which our framework outperforms standard smart predict-then-optimize and maintain its superiority even when the standard method is equipped with regularized upstream predictions. Numerical experiments validate that our robust framework consistently yields significant performance improvements over standard methods, both in out-of-sample terms and in training stability.
[LG-42] Prompt as a Data Type: In-Database LLM Prompt Management and Rewriting
链接: https://arxiv.org/abs/2607.21756
作者: Denis Mayr Lima Martins,Gottfried Vossen
类目: Databases (cs.DB); Machine Learning (cs.LG)
*备注:
Abstract:Large Language Models (LLMs) are increasingly used in database-backed applications to classify tuples, filter records using semantic predicates, extract structured attributes, and enrich query results. Yet the prompt that start these computations are typically stored outside the DBMS in unstructured formats, making them invisible to query execution, metadata management, and optimization. Drawing on Stonebraker’s QUEL as a Data Type and the principles of reflective programming, this paper introduces PromptDB, a database system that treats prompts as tuple-level database values. PromptDB provides a logical PROMPT datatype whose values store a template, bindings to tuple attributes, model metadata, and task metadata. Relations may contain PROMPT attributes directly in base tables, or expose them through views over joined tuples. Users query prompt-valued attributes through generated evaluation views, while the system internally renders, rewrites, optimizes, and executes prompts through an EVAL operator. Making prompts database-visible creates a new optimization space. The key idea is to bring query-optimizer thinking to prompts: just as query optimizers exploit database metadata to rewrite SQL plans, PromptDB exploits database metadata to rewrite prompts. We evaluate PromptDB on synthetic and real-world data workloads across different tasks. The results show how database-guided rewriting improves output validity and yields favorable cost-quality trade-offs compared with static, manually written prompts.
[LG-43] Parameter-free Adaptive Sparse Attention via Compression-Based Content Selection
链接: https://arxiv.org/abs/2607.21752
作者: Debarshi Kundu,Swaroop Ghosh,Vasant Honavar
类目: Machine Learning (cs.LG)
*备注:
Abstract:Data-adaptive sparse attention masks substantially outperform fixed patterns (e.g., BigBird and Longformer) and can even exceed dense attention on long sequences. Existing adaptive approaches—including SBM-Transformer, Dynamic Mask Attention, and NSA—typically require additional learnable parameters, custom gradient estimators, or specialized CUDA kernels. We show that classical data compression provides an effective masking signal with \textbfno additional parameters. By computing per-block gzip compression ratios, we identify non-redundant content blocks and route long-range attention selectively through them. Intuitively, blocks that gzip cannot compress contain information not predictable from local repetition, making them natural long-range attention targets. Because the compression profile is input-dependent, the resulting sparse mask adapts dynamically to content without learned parameters, auxiliary losses, or custom kernels. On PG-19 byte-level language modeling at 92M parameters with 8K context, our method achieves 1.71 bits-per-byte (BPB), outperforming dense attention (2.89), BigBird (2.34), Longformer (3.21), and a reimplemented SBM-Transformer (3.38)—the only learned-mask baseline—by up to 1.67 BPB while adding no parameters. The advantage grows with sequence length, with the gap over BigBird widening from 0.05 BPB at 4K context to 0.63 BPB at 8K, while convergence is 3.3 \times faster. Subjects: Machine Learning (cs.LG) Cite as: arXiv:2607.21752 [cs.LG] (or arXiv:2607.21752v1 [cs.LG] for this version) https://doi.org/10.48550/arXiv.2607.21752 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[LG-44] RED-PIM: Reducing Data Movement for Transformers using Processing-in-Memory
链接: https://arxiv.org/abs/2607.21731
作者: Zahra Yousefijamarani,Alaa Alameldeen
类目: Machine Learning (cs.LG); Hardware Architecture (cs.AR)
*备注:
Abstract:Transformers are widely used across many domains, including natural language processing, computer vision, web search, and DNA sequence analysis. Given their broad applicability, improving the performance of transformer models is critical. However, the high volume of data movement between processing units and memory during attention operations significantly limits their efficiency. Processing-In-Memory (PIM) mitigates this issue by performing computations directly inside memory. While prior work has proposed PIM-based transformer implementations, they suffer from costly inter-bank communication, and struggle to scale due to the limited capacity of memory banks. As a result, attention-related data must be split across banks, diminishing the potential benefits of PIM. In this work, we propose RED-PIM, an algorithm-architecture co-design that reduces attention latency by minimizing inter-bank data movement from O(N^2) to O(N) and shrinking intermediate attention matrices from N x N to d x d. By reorganizing matrix operations, performing computations locally, and employing an optimized data transfer strategy, RED-PIM significantly reduces computation cost and interconnect traffic. Compared to baseline PIM implementation, RED-PIM achieves inference time reductions ranging from 16.05% to 99.99% (geometric mean of 66.42%), with the largest gains on longer sequences. On real-world datasets, RED-PIM improves performance by 99.60% for long documents and 13.44% for shorter ones, while maintaining or improving accuracy. These results demonstrate RED-PIM’s effectiveness for scalable and efficient transformer inference. Subjects: Machine Learning (cs.LG); Hardware Architecture (cs.AR) Cite as: arXiv:2607.21731 [cs.LG] (or arXiv:2607.21731v1 [cs.LG] for this version) https://doi.org/10.48550/arXiv.2607.21731 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[LG-45] An Introduction to Bayesian and Frequentist Simulation-Based Inference with Machine Learning
链接: https://arxiv.org/abs/2607.21702
作者: Maximilian Dax,Theo Heimel,Gilles Louppe
类目: Machine Learning (cs.LG); Cosmology and Nongalactic Astrophysics (astro-ph.CO); Astrophysics of Galaxies (astro-ph.GA); High Energy Physics - Experiment (hep-ex); High Energy Physics - Phenomenology (hep-ph); Machine Learning (stat.ML)
*备注: 39 pages, 13 figures
Abstract:Simulation-based inference (SBI) with machine learning is an increasingly important tool for solving inverse problems in science and engineering, including parameter inference and the inversion of detector effects. We provide an overview of the Bayesian and frequentist statistical frameworks, describe how machine-learning-based SBI methods, such as neural posterior estimation and neural likelihood estimation, can be used for parameter estimation within these frameworks, and show that the same methods can also be applied to Empirical Bayes or unfolding tasks. We also discuss how to validate inference results and the limitations of SBI with machine learning.
[LG-46] CARNet Cycle-Conditioned Core Aggregation and Redistribution for Multivariate Time Series Forecasting ECML KDD2026
链接: https://arxiv.org/abs/2607.21681
作者: Awsaf Tausif Adib,Md. Shahria Sarker Shuvo,Md. Estehaar Ahmed Emon,Mustafa Kamal,Fuad Rahman,Shafin Rahman,Nabeel Mohammed
类目: Machine Learning (cs.LG)
*备注: Accepted at ECML PKDD 2026 (Research Track). Shortlisted for Best Research Track Student Paper Award
Abstract:Accurately modeling cross-variate dependencies remains a key challenge in multivariate time series forecasting, particularly in the presence of strong periodic patterns. Many existing approaches rely on attention-based mechanisms that incur quadratic complexity and scale poorly with increasing numbers of variates. Recent attention-free aggregation models address this issue through linear-complexity core-based interactions, but they do not explicitly leverage the global periodic structure present in the data. To overcome this limitation, we propose CARNet, a Cycle-Conditioned Core Aggregation and Redistribution framework that integrates global recurrent cycle information into efficient core based interaction modeling via Multihead Core Aggregation. Extensive experiments on multiple real-world multivariate forecasting benchmarks demonstrate that CARNet consistently outperforms strong transformer and non-attention baselines across diverse prediction horizons while preserving linear-complexity modeling of cross-variate dependencies.
[LG-47] Encoding Invisible Causation for Bridge Diagnostic Agents : Triple-Guided Retrieval-Augmented Fine-Tuning with QLoRA
链接: https://arxiv.org/abs/2607.21680
作者: Takato Yasuno
类目: Machine Learning (cs.LG)
*备注: 13 pages, 7 figures, 6 tables
Abstract:Bridge infrastructure deteriorates gradually, yet its root causes—salt intrusion, freezing, fatigue cracking, and others—remain invisible to the naked eye. Expert diagnosis relies on tacit knowledge built over years of practice. We address the challenge of automating this latent causal reasoning by proposing a Damage Cause Encoder that classifies 10-class damage causes from visible damage descriptions S_i for use in autonomous bridge diagnostic agents. Our approach chains three components: (i)Knowledge Triple Extraction—a large language model extracts causal triples of the form (damage \xrightarrow\mathttcaused_by cause) from 15–35 diagnostic PDF manuals and indexes them in a FAISS vector store; (ii)Retrieval-Augmented Context—at training and inference time, relevant causal triples \mathcalC_i are retrieved and concatenated with S_i , converting implicit domain knowledge into explicit Encoder context; (iii)Systematic Fine-tuning Comparison—we conduct a rigorous comparison of LoRA, QLoRA, and QA-LoRA on a fixed Golden Testset (116 stratified samples), demonstrating that QLoRA achieves the optimal trade-off: identical test accuracy (87.07%) to full-precision LoRA, 11% faster inference, 72% lower GPU memory, and superior generalization across diverse unseen inputs. A controlled Golden Testset—stratified, deduplicated, and difficulty-tagged—is introduced as a reusable benchmark contribution. QLoRA further outperforms LoRA by 13 percentage points on a 100-sample diverse evaluation spanning all 10 damage cause this http URL findings enable memory-efficient, high-accuracy diagnostic agents on consumer-grade hardware for edge deployment.
[LG-48] Neural Feature Governance: Extending Atom Prevalence
链接: https://arxiv.org/abs/2607.21671
作者: Idris Karel Seunda Ekwe,Patrick Tenga Shako,Ernest Parfait Fokoué
类目: Machine Learning (cs.LG)
*备注: 32 pages, 4 figures
Abstract:Neural network compression and interpretability remain open challenges in modern deep learn- ing, where billion-parameter architectures deliver impressive accuracy at the cost of trans- parency, computational efficiency, and reliable uncertainty quantification. This paper introduces Neural Atom Prevalence (NAP), a principled Bayesian framework for structured node-level model selection in feedforward neural networks. NAP introduces the neural atom (activation unit) and functions as a hybrid method operating through a four-phase pipeline: Bayesian Lottery Ticket (BLT) identification via Iterative Magnitude Pruning (IMP), soft variational training of the Spike and Slab Independent Gaussian (SS-IG) model, Poisson-Binomial (PB) optimal layer-size selection, and Bayesian fine-tuning to produce a sparse, stable, interpretable, and accurate model. Extensive empirical validation across simulated nonlinear regression, two UCI benchmark datasets (Concrete, YearPredictionMSD), and the MNIST image classification task demonstrates that NAP achieves state-of-the-art structural sparsity, reducing active nodes to as few as 8% of the original dense architecture on MNIST, while well-calibrated probabilisti- cally: the aleatoric-epistemic uncertainty decomposition reveals that model ignorance accounts for only 3 to 4% of total predictive variance across all experiments, and regression reliability diagrams confirm a near-nominal predictive interval coverage (93.4% observed against a 95% target). These results establish NAP as a reliable, theoretically grounded, and computation- ally tractable solution to the simultaneous pursuit of sparsity, accuracy, interpretability, and uncertainty quantification in Bayesian neural networks.
[LG-49] Physically Constrained Federated Additive Models for O-RAN SLA-Risk Prediction
链接: https://arxiv.org/abs/2607.21665
作者: Aubida A. Al-Hameed,Mohammed M. H. Qazzaz,Maryam Hafeez,Syed A. Zaidi
类目: Machine Learning (cs.LG); Systems and Control (eess.SY)
*备注: ACM conference 6 pages
Abstract:Proactive service assurance in O-RAN requires predicting per-slice SLA violations before they occur. The prediction model must be auditable by operators and must train across base stations without pooling per-slice KPIs, which are commercially sensitive because slices are leased to individual tenants. Neural additive models (NAMs) offer auditability because each KPI contributes through a visible shape function. However, visibility alone does not guarantee physical validity. On the ColO-RAN testbed dataset, unconstrained NAMs learn effects that contradict wireless physics, for example predicting higher risk when channel quality improves. This failure appears under both local and centralized training, and non-IID federated averaging worsens it. We present Monotone FedNAM, a federated additive model in which KPIs with unambiguous physical direction are represented as monotone splines whose constraints survive FedAvg aggregation by construction, while contestable KPIs remain unconstrained. The model trains and operates as a Non-RT RIC rApp and is compact enough for deployment as a Near-RT RIC xApp. Monotone FedNAM eliminates all monotonicity violations, raises constrained shape consistency from 0.71 to 1.00, generalizes to an unseen scheduling policy, and reduces uplink traffic by 65%, at a cost of 0.04 to 0.07 AUC. These results show that physically constrained federated additive models can support auditable SLA risk inference for multi-tenant O-RAN service assurance
[LG-50] Shallower ReLU Network Representations via Exact Linear Algebra
链接: https://arxiv.org/abs/2607.21651
作者: Kilian Rueß,Gennadiy Averkov,Florestan Brunck,Moritz Grillo,Christoph Hertrich,Georg Loho,Jack Stade,Moritz Stargalla,Matthew Sun,Martin Winter
类目: Machine Learning (cs.LG); Neural and Evolutionary Computing (cs.NE); Combinatorics (math.CO)
*备注:
Abstract:We prove that the maximum of n real numbers is exactly representable by a ReLU network with two hidden layers for every n\le 10 . The constructions are obtained by reducing the problem to exact rational linear algebra: after a symmetry reduction, the necessary cancellations are encoded in finite linear systems over \mathbbQ , which we solve and verify computationally. The representation of \max_10 has a structured first hidden layer consisting only of pairwise maxima, a feature that allows it to be recursively substituted into larger networks. We use this to show that for every n10 , the maximum \max_n can be exactly represented with \lceil\log_5 (n / 2)\rceil+1 \log_5(n) +1.5694 hidden layers. Via the generalized hinging-hyperplane representation [Wang, Sun, IEEE Trans. Inf. Theory 2005], the same depth bound holds for all continuous piecewise-linear functions on \mathbbR^d , with d+1 in place of n . In particular, every continuous piecewise-linear function on \mathbbR^d for d\le 9 admits a two-hidden-layer ReLU representation. Our results improve on [Bakaev, Brunck, Hertrich, Stade, Yehudayoff, STOC’26]. In that work, the authors established a two-hidden-layer representation for \max_5 and an upper bound of \lceil\log_3 (n-2)\rceil+1 hidden layers for \max_n .
[LG-51] Adjustment Speed as a Safety Constraint for Nonstationary Reinforcement Learning
链接: https://arxiv.org/abs/2607.21646
作者: Timothy Tomashevskiy
类目: Machine Learning (cs.LG)
*备注: 15 pages, 5 figures, 2 tables. Preprint
Abstract:Ensuring safety in reinforcement learning under nonstationarity requires determining whether a learning system can safely adapt to forecasted environmental change within the required recovery horizon. Existing safe reinforcement learning methods typically assume stationary environments and do not explicitly consider adaptation speed as a safety concern. However, when environments evolve over time, delayed adaptation may result in transient unsafe behavior. This paper proposes adjustment speed as a safety constraint for nonstationary reinforcement learning. The central idea is to define safety in terms of adaptation feasibility: future states or regions may become unsafe when the adaptation required to remain safe exceeds the learning system’s calibrated recovery capacity. The proposed framework uses learned context representations and short-horizon context forecasts to estimate adaptation demand and compare it with the agent’s achievable adaptation capacity. When predicted adaptation demand exceeds the calibrated recovery capacity, the framework proactively tightens the admissible action set and activates an action-level shield to reduce unsafe behavior before violations occur. Experiments in a nonstationary driving environment show that the proposed approach primarily reduces safety violations in short-horizon windows aligned with context changes. Ablation studies further show that shielding is more conservative for peak- and tail-risk suppression, while optimization-level adjustment provides additional reductions in short-horizon switch-conditioned violations. These results support adaptation feasibility as a practical safety principle for reinforcement learning under nonstationarity and demonstrate that proactive intervention can improve safety during periods of environmental change. Comments: 15 pages, 5 figures, 2 tables. Preprint Subjects: Machine Learning (cs.LG) Cite as: arXiv:2607.21646 [cs.LG] (or arXiv:2607.21646v1 [cs.LG] for this version) https://doi.org/10.48550/arXiv.2607.21646 Focus to learn more arXiv-issued DOI via DataCite Submission history From: Timofey Tomashevskiy [view email] [v1] Tue, 21 Jul 2026 23:52:07 UTC (405 KB) Full-text links: Access Paper: View a PDF of the paper titled Adjustment Speed as a Safety Constraint for Nonstationary Reinforcement Learning, by Timothy TomashevskiyView PDFHTML (experimental)TeX Source view license Current browse context: cs.LG prev | next new | recent | 2026-07 Change to browse by: cs References Citations NASA ADSGoogle Scholar Semantic Scholar export BibTeX citation Loading… BibTeX formatted citation loading… Data provided by: Bookmark checked="checked"class=“labs-tab-input”> Bibliographic Tools Bibliographic and Citation Tools Bibliographic Explorer Toggle Bibliographic Explorer (What is the Explorer?) Connected Papers Toggle Connected Papers (What is Connected Papers?) Litmaps Toggle Litmaps (What is Litmaps?) scite.ai Toggle scite Smart Citations (What are Smart Citations?) Code, Data, Media Code, Data and Media Associated with this Article alphaXiv Toggle alphaXiv (What is alphaXiv?) Links to Code Toggle CatalyzeX Code Finder for Papers (What is CatalyzeX?) DagsHub Toggle DagsHub (What is DagsHub?) GotitPub Toggle Gotit.pub (What is GotitPub?) Huggingface Toggle Hugging Face (What is Huggingface?) ScienceCast Toggle ScienceCast (What is ScienceCast?) Demos Demos Replicate Toggle Replicate (What is Replicate?) Spaces Toggle Hugging Face Spaces (What is Spaces?) Spaces Toggle TXYZ.AI (What is TXYZ.AI?) Related Papers Recommenders and Search Tools Link to Influence Flower Influence Flower (What are Influence Flowers?) Core recommender toggle CORE Recommender (What is CORE?) IArxiv recommender toggle IArxiv Recommender (What is IArxiv?) Author Venue Institution Topic About arXivLabs arXivLabs: experimental projects with community collaborators arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website. Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them. Have an idea for a project that will add value for arXiv’s community? Learn more about arXivLabs. Which authors of this paper are endorsers? | Disable MathJax (What is MathJax?) mathjaxToggle(); We gratefully acknowledge support from our major funders, member institutions, , and all contributors. About Help Contact Subscribe Copyright Privacy Accessibility Operational Status (opens in new tab) Major funding support from
[LG-52] oward Goal-Agnostic Joint-Embedding Predictive Control of Partial Differential Equations
链接: https://arxiv.org/abs/2607.21644
作者: Jonathan Gallagher,Roberto Guglielmi
类目: Machine Learning (cs.LG); Systems and Control (eess.SY)
*备注: Associated code will be open sourced alongside the V2 submission
Abstract:We present a goal-agnostic control framework for partial differential equations (PDEs) built around a joint-embedding predictive architecture (JEPA). The small 2D ViT encoder and action-conditioned latent dynamics are trained offline without a reward or downstream goal, frozen, and reused by a model-predictive path integral (MPPI) controller. We find that when available, the control objective is better applied to an explicit physical observable (provided injectivity) than to minimizing raw Euclidean distance ( L^2 ) in the learned latent space. For a learned linear kinetic-energy (KE) probe on frozen latent rollouts we can reproduce held-out trajectories with R^2=0.989 , while requiring no change to the underlying world model. On the PDE Control Gym 2D Navier–Stokes benchmark, using KE-probe planning improves the matched 50-episode native reward from -12.08\pm0.86 for latent- L^2 planning to -10.90\pm0.91 (95% CI), while lowering last-quarter velocity-field RMSE from 0.0765 to 0.0692 . Across three intentionally withheld, dissimilar, aperiodic targets, KE planning lowers late field RMSE by 53% relative to latent- L^2 planning ( 0.0220 versus 0.0469 ), winning all 30 paired episodes. The same frozen model also supports controls targeting stabilization around a steady configuration via direct regulation of KE achieving 2.7% mean relative error. While the latent probe is brittle to measurement noise and missing pixels, we believe the results support the claim that latent dynamics can remain both dynamic and goal-agnostic while calibrated observables (granted they guarantee unique continuation) may be a better objective for state control
[LG-53] Quasi-Monte Carlo Initialization for Meta-Reinforcement Learning
链接: https://arxiv.org/abs/2607.21637
作者: Julian G. Soltes
类目: Machine Learning (cs.LG); Optimization and Control (math.OC)
*备注: 6 pages, 6 figures, 1 table
Abstract:This paper explores the efficacy of quasi-Monte Carlo (QMC) weight initialization for meta-reinforcement learning within modern benchmark environments. Various sampling methods are used to bound a population-based search and aggregate an optimal prior from a baseline set of tasks. The QMC meta-priors show improvements in training convergence compared to modern orthogonal (SB3) defaults when extrapolated to similar unseen continuous control environments. In dissimilar tasks, the orthogonal orientation was globally superior for an unbiased search.
[LG-54] oward User-Conditioned Evaluation of Personal LLM Agents under Temporal Interventions KDD KDD2026
链接: https://arxiv.org/abs/2607.21635
作者: Pin Qian,Su Wang,Yihang Chen,Qiaolin Yu,Xiaoyuan Wang,Zhitong Guo,Zhicheng Wang,Junxian You
类目: Machine Learning (cs.LG)
*备注: 9 pages, 2 figures, and 8 tables. Accepted for oral presentation at the ACM SIGKDD KDD 2026 Workshop on Personal Intelligence in the Agentic AI Era (PILA 2026)
Abstract:Personal agents maintain memories, learned skills, tool configurations, and policy state that evolve with each user. Existing agent benchmarks often evaluate these capabilities in isolation: tool benchmarks test invocation under fixed APIs, memory benchmarks test recall or forgetting, and safety benchmarks test static policy compliance. We argue that personal-agent evaluation requires a different protocol: replaying the same temporal intervention across different persistent user-conditioned states and measuring how failures propagate across agent components. We formalize this requirement as four conditions: explicit temporal intervention, persistent state across the intervention, induced cross-dimensional effects, and variation in user-conditioned state. A focused audit of public benchmark protocols selected by explicit inclusion criteria identifies several close cases. Under our explicitly narrow operationalization, we did not find a protocol in that audited set satisfying all four conditions. This claim is scoped as a focused gap analysis with bounded literature coverage. This position paper proposes a minimal benchmark design and candidate reporting metrics for user-conditioned adaptation. The result is a concrete design requirement for future personal-agent evaluation, with metrics used as reporting tools for that requirement.
[LG-55] Cloud-Native Evaluation-as-a-Service: A Microservices Architecture for Scalable AI Monitoring with Conformal Guarantees
链接: https://arxiv.org/abs/2607.21623
作者: Lei Yang
类目: Machine Learning (cs.LG)
*备注: 23 pages, 15 figures, 12 tables
Abstract:We present EaaS, a cloud-native reference architecture that operationalizes AI evaluation methods as six stateless Kubernetes microservices: conformal prediction with finite-sample-corrected Adaptive Prediction Sets, calibration assessment, drift detection via RFF-approximated Maximum Mean Discrepancy, fairness monitoring with bootstrap confidence intervals, a DAG-based pipeline orchestrator, and a result storage API. We validate four key methodological concerns. First, empirical coverage is consistent with the marginal conformal guarantee across K=50 random calibration/test splits, with mean coverage within 1.4 percentage points of the nominal target. Second, all four MMLU answer tokens appear in the top-20 logprobs with 0% imputation needed, and simulated imputation at 10% produces less than 1.5% coverage impact. Third, RFF-MMD achieves 100% detection power for mild and severe drift at the median heuristic bandwidth, with Type I error between 5-8.5%. Fourth, fairness monitoring on the UCI Adult Income dataset reveals significant demographic parity disparities by race (DP gap=0.33) with stable alerts across sequential batches. Conformal prediction and calibration services achieve sub-2ms p99 latency at batch size 100; RFF-MMD requires ~500ms suited for periodic batch monitoring. A comparison with four open-source tools suggests that, to the best of our knowledge, no current platform combines conformal-prediction-as-a-service, microservice decomposition, and DAG-based orchestration.
[LG-56] Graph-Based Correlation Matrix Generation: A Convex Optimization Approach
链接: https://arxiv.org/abs/2607.22436
作者: Ali Fakhar(UGA),K{é}vin Polisano(UGA),Ir{è}ne Gannaz(G-SCOP_GROG, G-SCOP, Grenoble INP, UGA),Sophie Achard(STATIFY, LJK, UGA)
类目: Machine Learning (stat.ML); Machine Learning (cs.LG)
*备注:
Abstract:This work addresses the generation of theoretical correlation matrices with prescribed sparsity patterns associated to graph structures. We propose a novel convex optimization framework in which an initial matrix is projected onto an elliptope under a positive semidefiniteness constraint. Several numerical schemes are implemented and compared. The problem falls within the broader class of matrix completion, where off-diagonal entries corresponding to absent edges are fixed to zero and diagonal entries are fixed to one. Beyond this structural constraint, the approach offers greater flexibility than existing methods by allowing control over the mean of the off-diagonal entry distribution, enabling the generation of correlation matrices that better reflect realistic data. This procedure is not designed to yield a uniform distribution over the feasible set; rather, it provides a principled and tunable way to construct correlation matrices suitable for benchmarking statistical methods for graphical model inference. Theoretical guarantees on the existence of solutions are established, both in the general setting and under the additional mean constraint. Simulation studies illustrate the properties of the generated matrices with respect to graph structure. The methodology is applied to two real-world datasets from neuroscience and finance, and a comparison with GAN-based correlation matrix generation is provided.
[LG-57] Learning Ergodic Dynamical Systems from a Finite Trajectory
链接: https://arxiv.org/abs/2607.22399
作者: Oleksii Kachaiev,Silvia Villa,Lorenzo Rosasco
类目: Machine Learning (stat.ML); Machine Learning (cs.LG)
*备注:
Abstract:We consider the problem of learning from a single finite trajectory of an ergodic stochastic dynamical system. More precisely, we study discrete-time autonomous stochastic systems defining time-homogeneous Markov processes. We first focus on estimating the optimal one-step prediction function by nonlinear least squares, and derive high-probability guarantees measured with respect to the invariant measure of the process. These results make explicit how the non-independent and non-identically distributed nature of trajectory data modifies the classical statistical learning analysis. We then extend the framework to higher-order systems and finite-state spaces. Finally, we show that the same least squares and concentration arguments naturally extend to learning Koopman operators. Our approach combines tools from statistical learning theory and quantitative ergodic theory for Markov chains. It relies, in particular, on a concentration inequality for Hilbert-space-valued additive functionals of uniformly geometrically ergodic Markov chains.
[LG-58] Learning Bidirectional Causal Interactions with Heteroscedastic Neural Networks
链接: https://arxiv.org/abs/2607.22313
作者: Masahiro Tanaka
类目: Machine Learning (stat.ML); Machine Learning (cs.LG); Methodology (stat.ME)
*备注:
Abstract:Estimating contemporaneous bidirectional interactions from observational data is difficult because each outcome is endogenous to the other, while flexible regressions may capture only reduced-form dependence. This paper proposes SEM-DNN, a heteroscedastic neural simultaneous-equation estimator that learns reciprocal structural interactions without external instruments. Identification exploits conditional covariance diagonalization: when structural shocks have zero conditional means, are conditionally uncorrelated given predetermined covariates, and exhibit nonproportional conditional variances, only the true interaction coefficients diagonalize the conditional residual covariance across the feature space. The method jointly approximates nonlinear structural mean functions and feature-dependent variances using a diagonal Gaussian quasi-likelihood that incorporates the simultaneous-system Jacobian. We establish unique identification and positive-definite local curvature of the profiled population criterion and show that, under neural-profile compatibility conditions, the implemented neural criterion inherits this curvature despite nonunique network parameterizations. The coefficients admit a causal interpretation when the structural equations represent autonomous mechanisms that remain invariant under the relevant interventions. Monte Carlo experiments with nonlinear, high-dimensional nuisance functions and non-Gaussian shocks show that SEM-DNN recovers structural effects more reliably than parametric, kernel-based, and separate-equation neural alternatives as information increases, although at greater computational cost. An application to ready-to-eat cereal scanner data illustrates how the method can study contemporaneous price-sales feedback and assess identification strength, residual diagonalization, variance calibration, and optimization sensitivity.
[LG-59] Hopformer: Homogeneity-Pursuit Transformer for Time Series Forecasting
链接: https://arxiv.org/abs/2607.22299
作者: Wan Zhang,Qinjie Lin,Chan Lee,Weijian Li,Han Liu,Kai Zhang
类目: Machine Learning (stat.ML); Machine Learning (cs.LG)
*备注:
Abstract:Forecasting multiple time-series with high-dimensional covariates presents a core challenge: unifying common temporal patterns while retaining meaningful series-specific information. We introduce Hopformer (Homogeneity-Pursuit Transformer), a two-stage framework that addresses this challenge. In the first stage, we perform a Sparsity Pattern Aggregation (SPA) scheme extracting a common low-variance trend that incorporates the covariates. This acts as a homogenization layer. In the second stage, a LoRA-fine-tuned Transformer models the remaining complex dependencies in the residual. Our method is theoretically grounded. We prove that SPA achieves a near-optimal bias-variance trade-off via an oracle inequality. We also provide generalization bounds for the second stage under dependent time series data. Hopformer sets a new state of the art, improving MASE by an average of 6.56% across synthetic and real-world forecasting benchmarks.
[LG-60] General Value Functions for Remaining Useful Life and Failure-Mode Prediction
链接: https://arxiv.org/abs/2607.22268
作者: Hao Yan,Ali Sarabi,Qing Zou,Boyang Xu
类目: Machine Learning (stat.ML); Machine Learning (cs.LG); Applications (stat.AP)
*备注:
Abstract:Remaining useful life (RUL) prediction and failure-mode classification are central tasks in predictive maintenance. Many data-driven pipelines use fixed-window supervised learning with complete terminal labels; such routes do not naturally encode the temporal recursion linking successive degradation-state predictions when observations are partial or unit identities are unavailable. We formulate prognostics as vector General Value Function (GVF) prediction on an absorbing degradation process, treating RUL and failure-mode probabilities as temporally consistent targets rather than independent window-level labels, and estimate them with a multi-step temporal-difference estimator, TD( n,\lambda ). Supporting theory identifies the Bellman fixed point of the vector GVFs, characterizes the linear projected-TD limit and its relation to complete-return Monte Carlo regression under realizability, and explains when bootstrapped TD targets are less variable than Monte Carlo returns. On an event-triggered multimode simulation and NASA C-MAPSS label-scarce stitch data, TD improves RUL and failure-mode prediction relative to a supervised same-backbone Monte Carlo control, especially under scarce complete labels. Practically, fragmented, identity-free degradation records can contribute local Bellman transitions instead of being discarded until complete run-to-failure labels are available.
[LG-61] Variational Low-rank Tensor Decomposition for Multisubject Spatiotemporal Data Analysis
链接: https://arxiv.org/abs/2607.22262
作者: Laura M. Montaldo,Ricardo A. Borsoi,Sebastian Miron,Tulay Adali
类目: Machine Learning (stat.ML); Machine Learning (cs.LG); Signal Processing (eess.SP)
*备注: 6 pages, 2 figures, conference
Abstract:Modeling shared and subject-specific structure in multisubject spatiotemporal data remains challenging, particularly in neuroimaging, where both spatial and temporal patterns exhibit rich variability across subjects. Existing matrix and tensor decompositions provide interpretable factorizations, but rely on fixed multilinear structures or coupling schemes that may limit their flexibility in capturing complex variability. In this work, we introduce a spatiotemporal variational tensor decomposition (ST-VTD) framework that combines a tensor factorization generative model with structured priors to jointly represent spatial maps and temporal dynamics. Spatial factors are regularized to promote a low-rank structure inspired by the LL1 decomposition, while temporal factors are modeled using a learned Long short-term memory (LSTM)-based prior, enabling flexible and adaptive dynamics. Posterior inference is performed using an amortized variational formulation by unrolling iterations of an optimization algorithm, leading to an interpretable and parameter-efficient architecture. The proposed inference framework employs a warm-start strategy based on group independent component analysis, which we found to improve optimization performance. Experiments on a realistic synthetic functional MRI (fMRI) dataset demonstrate that the proposed approach significantly improves latent factor recovery compared with representative classical and probabilistic decomposition benchmarks.
[LG-62] Convergence analysis of a family of Zermelo-type iterations for the Bradley–Terry model
链接: https://arxiv.org/abs/2607.22221
作者: Ruijian Han,Ding Lu,Yiming Xu
类目: Machine Learning (stat.ML); Machine Learning (cs.LG); Numerical Analysis (math.NA); Computation (stat.CO)
*备注: 46 pages
Abstract:Zermelo’s algorithm is a classical method for computing the maximum likelihood estimator in the Bradley–Terry (BT) model, but its convergence can be slow in practice. To accelerate computation, Newman introduced a family of Zermelo-type fixed-point iterations parameterized by \alpha , with Zermelo’s algorithm recovered at \alpha=1 . Empirical evidence suggests that the choice \alpha=0 often converges substantially faster, making it a promising alternative, yet the mechanism underlying this acceleration remains elusive. This paper provides theoretical insight into this phenomenon through a systematic local convergence analysis. We derive closed-form expressions for local convergence factors under synchronous and asynchronous updates and analyze their dependence on \alpha via spectral analysis of the associated Jacobian matrices. For synchronous updates, we show that the algorithm may fail to converge when \alpha1 , and its local convergence factor is quasi-convex in \alpha under the population BT model. In contrast, asynchronous updates are always locally convergent, and their local convergence factor is provably monotonically increasing in \alpha under the population BT model of consistently ordered bipartite comparison graphs, establishing the optimality of \alpha=0 in this setting. We further establish asymptotic approximation results for the population convergence factors under the BT model, justifying their practical relevance. Numerical experiments on synthetic and real-world datasets confirm the theory. Our analysis complements existing convergence results and shows that the acceleration of \alpha=0 arises not only from the parameter choice but, more importantly, from the use of asynchronous updates.
[LG-63] QC-PHAST Search: Classical–Quantum Query Benchmarks for Finite-Pool Rare-Regime Discovery
链接: https://arxiv.org/abs/2607.21995
作者: Harsh Milind Tirhekar,Chandrajit Bajaj
类目: Quantum Physics (quant-ph); Emerging Technologies (cs.ET); Machine Learning (cs.LG)
*备注: 70 pages, 16 figures, 36 tables
Abstract:Rare-regime discovery in parameterized dynamical systems is an active-search problem: find one verified parameter at which a scientifically defined qualitative threshold is crossed, even when acceptable candidates are rare, nonconvex, or fragmented. We introduce Quantum-Classical Phase-space and Stability-Threshold Search (QC-PHAST), an evidence-gated decision protocol and query-accounting framework for finite candidate libraries. A candidate induces a dynamical object, simulator-derived criticality score, and verified first-hit predicate. Scientific metadata and charged pilot evidence are used to assess whether equation-aware search, scalar-score active search, predicate-only search, or only a query-model comparison is admissible. The quantum row is the inherited Grover/Boyer–Brassard–Hoyer–Tapp (BBHT) unknown- M marked-set query reference; it is not a new quantum-search theorem, materialized circuit, or hardware-speedup claim. The result is a regime map. Direct boundary constructions, geometry controls, online simulator loops, and learned-label accounting further identify when classical structure, false positives, calibration cost, or state preparation erases the query-model margin. QC-PHAST is therefore an auditable protocol for deciding when a finite-pool marked-set reference is informative and when classical or resource-aware search should dominate.
[LG-64] Efficient Online LLM Watermark Detection via Rao-Blackwellized E-Processes
链接: https://arxiv.org/abs/2607.21958
作者: Lu Luo,Dandan Mo,Chengdong Xu,Ting Li,Jinhan Xie,Huiqiong Li,Niansheng Tang
类目: Machine Learning (stat.ML); Machine Learning (cs.LG)
*备注: 25 pages, 3 figures
Abstract:As large language models (LLMs) are increasingly deployed, reliable and efficient mechanisms for distinguishing AI-generated text from human-written content have become essential. Statistical watermarking has emerged as a promising solution, yet most existing methods are typically fixed-horizon procedures, precluding valid early stopping in streaming generation. In this paper, we develop an efficient online watermark detection framework with anytime-valid inference based on Rao-Blackwellized e-processes, enabling recursive token-level evidence updates without storing the full history. In particular, we instantiate the framework for the Gumbel-max watermark and reduce the original token-level dependence testing problem to a pivot-induced sequential testing problem with an explicit null distribution. Theoretically, we prove anytime-valid Type I error control under arbitrary optional stopping and establish positive asymptotic log-growth under watermarking, implying consistency of the proposed stopping rules. Simulations and experiments on real LLM-generated text demonstrate efficient online detection with rigorous anytime-valid guarantees.
[LG-65] Distributional Determinantal Point Process for Repulsive Clustering of Distributions
链接: https://arxiv.org/abs/2607.21847
作者: Khai Nguyen,Yang Ni,Elizabeth Juarez-Colunga,Peter Mueller
类目: Methodology (stat.ME); Machine Learning (cs.LG); Applications (stat.AP); Computation (stat.CO); Machine Learning (stat.ML)
*备注: 60 pages, 15 figures
Abstract:We introduce the distributional determinantal point process (dDPP) as a novel repulsive point process whose atoms are probability distributions rather than points in a real space. The dDPP is constructed via an L-ensemble with a sliced Wasserstein (SW) kernel between distributions. We show its validity as a well-defined point process. In the discrete setting, we derive concentration results for plug-in estimators of the L-ensemble, the correlation kernel, and their determinants given i.i.d. samples from the distributional atoms. Leveraging this framework, we propose a distribution-valued random partition model by way of a repulsive generalized Bayesian mixture model. The model places a dDPP prior over the atoms of the mixing measure and defines a generalized likelihood based on SW distance. To summarize posterior inference, we develop a decision-theoretic approach to report a point estimate of the mixing measure as a Bayes rule under a hierarchical optimal transport utility function. The latter is a natural choice given that the mixing measure is itself a distribution over distributions. We use the proposed framework for inference with single-cell gene expression data and human epilepsy data, producing interpretable and well-separated clusters that reflect meaningful structure in the data.
[LG-66] Simulation-Based Empirical Bayes
链接: https://arxiv.org/abs/2607.21843
作者: Xinwei Shen,Diana Cai,Cheng Zhang,David M. Blei
类目: Machine Learning (stat.ML); Machine Learning (cs.LG)
*备注:
Abstract:Empirical Bayes (EB) performs simultaneous inference across many related latent variables. Classical EB assumes that the likelihood p(x | z) is tractable. In many scientific applications, however, the likelihood is available only through a simulator. This paper develops EB for such implicit likelihoods. We introduce simulation-based empirical Bayes (SBEB), which connects nonparametric EB to simulation-based inference (SBI). SBEB computes EB estimates without an explicit density by using the observed data, simulator samples, and an amortized inference network. SBEB iteratively refines the fitted EB prior toward the population prior. With several scientific simulators and real-world data, we demonstrate that SBEB improves accuracy over SBI with a fixed prior.
[LG-67] Longitudinal Random Forests for Sparse and Irregular Response Trajectories
链接: https://arxiv.org/abs/2607.21817
作者: Yangsheng Wang,Xiaotian Dai,Haoda Fu,Guifang Fu
类目: Methodology (stat.ME); Machine Learning (cs.LG); Machine Learning (stat.ML)
*备注:
Abstract:Longitudinal studies often collect data at sparse, irregular, and unequally spaced time points. Such heterogeneity is often driven by subject-specific covariates, yet existing methods have been restricted to a scalar endpoint value, completely neglecting the underlying response trajectories. We propose a novel Longitudinal Random Forest (LRF) framework that leverages tree-based ensemble machine learning with adaptive node-wise longitudinal trajectory estimation. The LRF framework makes five methodological contributions. it captures each subject’s individual response trajectory while simultaneously accommodating within-node correlation, between-node heterogeneity, and nonlinear and interactive covariate effects. It introduces a novel trajectory-based splitting criterion that maximizes trajectory separation while incorporating a size-weighted penalty; it provides two variants, Principal Analysis by Conditional Expectation (LRF-PACE) and adaptive linear mixed-effects models (LRF-adaptiveLMM), which employ nonparametric and semiparametric node-wise smoothers, respectively, while learning covariate effects in a data-driven manner. It provides a comprehensive interpretation of covariates using both the classical trajectory-based permutation variable importance measure (PVIM) and a newly proposed finite-way interaction frequency count, and it not only predicts entire trajectories for new subjects but also forecasts future trajectories for existing subjects. Extensive simulation studies demonstrate that LRF achieves superior performance over several competing methods, even under severe sparsity. The practical significance of the LRF framework lies in its ability to address five important clinical questions.
[LG-68] Natural Invariant Measures for Chaotic Game Dynamics: Finding Order in Chaos
链接: https://arxiv.org/abs/2607.21805
作者: Jakub Bielawski,Thiparat Chotibut,Fryderyk Falniowski,Michał Misiurewicz,Georgios Piliouras
类目: Dynamical Systems (math.DS); Machine Learning (cs.LG); Theoretical Economics (econ.TH)
*备注:
Abstract:We study the long-term behavior of the Multiplicative Weights Update (MWU) algorithm in game settings where learning dynamics frequently fail to converge to Nash equilibria and instead exhibit Li-Yorke chaos. While such chaos precludes the prediction of specific long-term strategy profiles, it does not imply a lack of statistical structure. We demonstrate that natural invariant measures - a fundamental concept from ergodic theory - provide the rigorous framework necessary to find order within this chaos. Focusing on a two-strategy congestion game, we prove that these measures allow for a comprehensive statistical characterization of the dynamics. Crucially, we show that this framework extends beyond simple strategy frequencies to \emphgeneral observables, enabling the precise calculation of long-term time averages for broad classes of economic metrics - including payoffs, social cost, and regret - despite chaos. Our results reveal that this simple learning algorithm captures the full spectrum of behaviors found in one-dimensional dynamical systems, from unique or multiple absolutely continuous invariant measures to complex periodic attractors as well as coexisting chaotic and stable (periodic) behaviors. By bridging game theory and dynamical systems, we show that statistical predictability is attainable even in the absence of pointwise convergence.
[LG-69] Reliability-Aware Bayesian Optimization of 1310 nm PCSELs with FDTD Verification
链接: https://arxiv.org/abs/2607.21772
作者: Jinglin Yu,Feiyang Wu,Longying Wen,Chongxian Yuan,Renjie Li,Zhaoyu Zhang
类目: Optics (physics.optics); Machine Learning (cs.LG); Applied Physics (physics.app-ph)
*备注:
Abstract:Near 1310 nm photonic-crystal surface-emitting lasers (PCSELs) are attractive narrow-beam sources for optical communication and sensing, but their final design refinement is costly. Small geometry changes simultaneously shift the band-edge resonance, cavity leakage, far-field divergence, and the numerical stability of a high- Q decay fit, while every full-wave trial requires a time-domain simulation. We couple a commercial finite-difference time-domain solver to a reliability-aware Bayesian optimization (BO) loop over eight local design variables. Each completed simulation updates the surrogate used to choose the next geometry. Candidate ranking combines wavelength and beam-quality requirements with a reliability-adjusted metric Q_\mathrmeff derived from the solver-reported relative fit-error estimate dQ/Q . Across three 80-evaluation runs from the same reference model, BO produced 5–15 candidates per run that passed the joint filter. Designs reconstructed from fresh model copies retained Q_\mathrmeff=4.33\times10^6 – 7.76\times10^6 , a 60–108-fold increase over the baseline metric, at 1308.23–1310.90~nm with approximately 0.84^\circ divergence. Under equal budgets, BO gave the highest mean strict-filter yield (9.0 candidates), compared with differential evolution (7.0) and Latin-hypercube sampling (1.5), although the controls occasionally matched the peak Q_\mathrmeff . Field maps, resonance spectra, and local perturbations further identify an index-related wavelength handle and a hole-size-related leakage handle. The resulting FDTD budget produces a pool of wavelength-compatible, narrow-beam, and reproducible high- Q PCSEL candidates without trusting a single optimistic decay fit.
[LG-70] Encoding orders and trees in real-valued functions
链接: https://arxiv.org/abs/2607.21761
作者: G Conant,C Terry
类目: Combinatorics (math.CO); Machine Learning (cs.LG); Logic (math.LO)
*备注: 27 pages
Abstract:We prove function-theoretic analogues of a quantitative result of Hodges on extracting the order property from a sufficiently large 2-tree coded in a binary relation. Similar analogues for functions were previously obtained by Daskalakis and Golowich and by Anderson and Benedikt. These results are from statistical learning theory, where 2-trees are captured by sequential fat-shattering dimension, and the order property is controlled by various notions of “thresholds”. Our first main result (Theorem 1.11) focuses on extracting a less restrictive kind of threshold from a tree, and yields significantly better bounds compared to what can be obtained from earlier results focusing on more restrictive versions. Part of the motivation for Theorem 1.11 lies in a companion paper, where this theorem is used to obtain efficient bounds in quantitative regularity lemmas for “stable functions”. Here will use Theorem 1.11 to reprove a result of Anderson and Benedikt in a stronger form and with improved bounds. We also use Theorem 1.11 to prove an at most double-exponential bound on dual sequential fat-shattering, which resolves an open problem. In our second main result (Theorem 1.14), we give a new proof of a result of Daskalakis and Golowich on extracting “tight thresholds” from large sequential fat-shattering dimension, with improved bounds. This resolves another open problem related to correcting the proof of a result claimed by Jung, Kim, and Tewari.
[LG-71] Prior laundering: learned priors with inherited undetectable overconfidence
链接: https://arxiv.org/abs/2607.21721
作者: Ali Siahkoohi,Sina Alemohammad
类目: Machine Learning (stat.ML); Machine Learning (cs.LG)
*备注:
Abstract:Learned generative priors are increasingly used for ill-posed Bayesian inverse problems, their posterior uncertainty treated as earned from data. But training one requires truths, scarce in seismic and medical imaging, so the recourse is an archive of legacy reconstructions—prior laundering. Where the measurements are uninformative the posterior reverts to the prior, so the uncertainty reported there is the archive’s, not the data’s, and nothing in deployment reveals it: truths differing only on those directions induce identical data laws, and self-consistency checks such as simulation-based calibration pass whatever the prior believes. That belief has an exact source: averaging the legacy posterior over the measurements yields the old regularizer advanced a single expectation–maximization step—improved where the data resolve, frozen where they cannot. It is overconfident wherever the inherited belief is tighter than the truth. A single-best archive is worse, collapsing the blind credible interval to zero width. Deployed, a diffusion prior fit to the archive under-covers the operator’s blind subspace, unlike a truth-trained control, and a normalizing flow does the same on a nonlinear groundwater operator. We recommend reporting which directions the measurements resolve, separating confidence the data support from belief inherited through the pipeline.
[LG-72] Explainable quantum-compressed machine learning for complex fluid flows
链接: https://arxiv.org/abs/2607.21688
作者: Xiao Xue,Maida Wang,Mingyang Gao,Minh Chung,Peter V. Coveney
类目: Fluid Dynamics (physics.flu-dyn); Machine Learning (cs.LG); Quantum Physics (quant-ph)
*备注:
Abstract:Machine-learning surrogates of physical systems face a paradox: explainable models facing the challenge of expressivity to capture complex nonlinear flows, whereas expressive deep surrogates match high-fidelity simulations only through massive parameterisations that turn the learned dynamics into a black box. Here, we introduce quantum-compressed machine learning (QCML), which resolves this tension by compressing the latent propagator of a flow surrogate from 524,288 trainable parameters to no more than 8 . This parameter reduction brings the learned dynamical law to the parameter scale of a physical constitutive relation rather than a black-box neural network, making the surrogate directly interpretable and controllable without sacrificing expressivity. The compression is realised by a structured quantum circuit whose unitary propagator constrains the latent spectrum to the unit circle exactly and by construction, replacing exponential error growth with linear accumulation over autoregressive rollouts. Classical regularisation only approximates this constraint: even a quantum-inspired classical baseline penalised towards unitarity collapses within one Lyapunov time on turbulent channel flow, whereas QCML remains stable over the full rollout. Shared phase and coupling angles parameterising the circuit correspond directly to modal frequencies and inter-mode interactions, giving the learned dynamics a physical interpretation in spectral space. On two patient-specific cardiovascular benchmarks, the structured QCML propagator matches the predictive accuracy of its classical counterpart on surface pressure spectra, pressure drop, and wall shear stress. These results establish QCML as a working component of scientific machine learning and a concrete contribution towards practical quantum advantage in real-world prediction.
[LG-73] Computer Vision Based Neurology Brain Activity Rejection Architecture and Implementation
链接: https://arxiv.org/abs/2607.21654
作者: Zag ElSayed,Nathan Suer,Grace Westerkamp,Jack Yanchen Liu,Makoto Miyakoshi,Craig Erickson,Ernest Pedapati
类目: Neurons and Cognition (q-bio.NC); Machine Learning (cs.LG); Data Analysis, Statistics and Probability (physics.data-an); Quantitative Methods (q-bio.QM)
*备注: 7 pages, 10 figures, Conference
Abstract:The electroencephalogram (EEG) is a valuable and widely applied tool for investigating brain disorders and behavioral changes. It offers a minimally restrictive and non-invasive method. However, challenges in using EEG for cognitive development studies include temporal resolution, signal source localization, and EEG artifacts. Careful consideration of these factors is essential for informed application of EEG technology. Independent component analysis (ICA) effectively isolates source generator processes from signals recorded by multiple, adjacent EEG scalp electrodes. Although ICA decomposition requires manual inspection, selection, and interpretation of independent components (ICs), this process is time consuming and demands expertise. Automated IC classification can achieve sufficient accuracy, expediting large scale EEG research and enabling near real time applications in conjunction with brain activity rejection tasks, which are crucial for medical specialists. This study introduces an automated computer vision based ICA rejection labeling tool compatible with widely used software interfaces like ICLabel and EEGLab. By automating the manual task, the proposed system reduces processing time by 7200 fold and achieves an accuracy of 89.45%.
附件下载
