cs.AI · 2026-10-17 · No. 146

Artificial Intelligence, 2026-10-17.

49 new papers in cs.AI. Titles, authors, abstracts. Links to arXiv. Want this in your inbox every morning? Subscribe →

01 — The papers

49 entries
  1. 01

    FDE-Gym: The Last Mile of AI Delivery

    Dongxu Yang, Hongda Yan, Renxin Wang, Baoli Zhang, Haojia Zuo, Ying Zhao, Youmin Chen, Xiaojian Liao, Congming Gao,...

    cs.AI · cs.CL · cs.SE

    Existing agent benchmarks typically evaluate agents on well-specified tasks defined by issue reports, specifications, or task tickets. Real-world customer delivery is different: forward-deployed engineers (FDEs) must clarify ambiguous requirements, work in live systems, and uncover unstated constraints. We introduce FDE-Gym, a benchmark of 95 end-to-end delivery tasks from 20 customer scenarios across 18 sectors, each adapted by industry...

    arxiv.org/abs/2610.22350 · PDF

  2. 02

    MLExpert: Towards a Generalist Agentic Model for Autonomous Machine Learning Engineering

    Shangheng Du, Jinxin Shi, Shuaiyu Zhang, Yiming Wu, Junpeng Zhao, Shiyang Feng, Xin Li, Liang He, Xiangchao Yan, Bo Zhang

    cs.AI · cs.CL · cs.LG

    Large language models (LLMs) are expanding from assisting scientific research and engineering to autonomously developing and improving AI systems, advancing AI for AI (AI4AI) as a pathway toward recursive self-improvement (RSI). Workflow-driven machine learning engineering (MLE) systems provide a practical path toward this pursuit through complex orchestration that manages long-horizon experimentation, while recent MLE systems adopt...

    arxiv.org/abs/2610.22344 · PDF

  3. 03

    Evaluation Awareness Beyond English: Do LLMs Act Differently When Tested in Indic Languages?

    Trizal Garg, Sibayan Mitra, Arya Karanjkar, Advait Singh, Jagat Sesh Challa, Yash Sinha, Siddharth Mehrotra, Dhruv...

    cs.AI · cs.CL

    Large language models (LLMs) can often tell when they are being tested. This ability, known as evaluation awareness, can lead models to act more cautiously during safety evaluations than in real use. Existing work has studied evaluation awareness almost entirely in English, although models serve users in many languages. Indic languages are spoken by over a billion people, yet evaluation awareness remains almost entirely unstudied in them. We...

    arxiv.org/abs/2610.22333 · PDF

  4. 04

    KnowGPO: Knowledge-Grouped Policy Optimization for Search Agents

    Siqi Chen, Zhengxi Lu, Yuchen Yan, Xingyu Wu, Qipeng Chen, Xin Zhang, Aiting Liu, Chao Deng, Jie Liu, Jin Ma,...

    cs.AI

    Deep search agents solve knowledge-intensive tasks through long-horizon, multi-turn interactions, where assigning credit to individual search decisions is fundamentally difficult. Outcome-based reinforcement learning provides only a single terminal reward shared across an entire trajectory. Recent process-supervision methods aim to assign credit at finer granularity, but they typically anchor credit to either surface-level trajectory states...

    arxiv.org/abs/2610.22331 · PDF

  5. 05

    Designing Transformers for Continuous Diffusion Language Models

    Weijian Luo, Yujian Chen, Hao Jiang, Guang Cheng, Tianyang Hu

    cs.AI

    Continuous diffusion language models (CDLMs) offer an alternative approach to parallel language generation; however, most existing studies directly use transformer architectures from either causal large language models (LLMs) or Diffusion Transformers (DiTs) for image generation, which were not optimized for CDLMs. In this paper, we conduct a systematic study on Transformer architectures for CDLMs and introduce Language Diffusion Transformer...

    arxiv.org/abs/2610.22330 · PDF

  6. 06

    A team of Gemini 4 Argon agents beats Minecraft

    Maxence Faldor, Satinder Singh, Roberta Raileanu

    cs.AI

    We placed four Gemini 4 Argon agents in a sandboxed Minecraft survival world and gave them one goal: defeat the Ender Dragon. Each agent controlled a player via an open-source library called Mineflayer and coordinated with its teammates through in-game chat and shared memory. No one assigned roles or wrote game logic. The agents planned, wrote, and debugged every program themselves. Researchers could watch the run on a live dashboard and...

    arxiv.org/abs/2610.22319 · PDF

  7. 07

    Private Retrieval Augmented Generation via Random Projection

    Dixi Yao, Tahseen Rabbani, Tian Li

    cs.AI · cs.CR

    Retrieval-Augmented Generation (RAG) enhances the capabilities of large language models (LLMs) by querying external structured knowledge. However, it can also introduce privacy risks by leaking sensitive information from the retrieval datastore. We propose a simple method to preserve datastore privacy in RAG systems via random projection. By applying the same projection to both datastore embeddings and query embeddings, our method provably...

    arxiv.org/abs/2610.22297 · PDF

  8. 08

    EliteReplay: Preserving Diverse Solutions in RL Post-Training via Quality-Diversity Search

    Hanane Nour Moussa, Doeun Lee, Sachin Kumar

    cs.AI

    Reinforcement learning has become a common post-training technique to improve the accuracy of LLMs on verifiable tasks. However, such training can narrow the set of solutions the model can produce. Several approaches have attempted to address this behavior-sharpening effect, but they often optimize metrics that do not reflect the diversity of correct solutions. Ideally, we want a model that can produce multiple good solutions, not only a...

    arxiv.org/abs/2610.22290 · PDF

  9. 09

    Improving Environment-Agnostic Coding Agents: The Role of Scaffolding

    Sergey Rodionov

    cs.AI

    Stronger language models may increasingly construct useful procedures for themselves, reducing the advantage of developer-supplied scaffolding. We investigate why this advantage persists and whether an environment-agnostic agent can be improved without environment-specific engineering. We evaluate a Codex-based agent in two nearly identical prompt variants on ARC-AGI-3, AGI Maze, and Terminal-Bench. Each variant is deployed with unchanged...

    arxiv.org/abs/2610.22289 · PDF

  10. 10

    LEASH: Uncertainty-Aware Weak-to-Strong Escalation and Oversight for Computer-Use Agents

    Qingni Wang, Changdae Oh, Yuzhe Yang, Sharon Li, Jiachen Yang, Ang Li, Xin Eric Wang

    cs.AI

    Computer-use agents increasingly tackle long-horizon tasks, where running a frontier model at every step is expensive and a small open-weight model alone often fails. Weak-to-strong escalation offers a middle ground where a weak agent acts by default and hands control to a strong one only when needed. The core challenge is deciding when to escalate, which requires an uncertainty signal cheap enough to check at every step. However, even after...

    arxiv.org/abs/2610.22280 · PDF

  11. 11

    X-RIC: An Agentic RAN-Intelligent Controller Benchmark for Enhancing Reinforcement Learning

    Adhwaa Al-Chaab, Ayman Younis, Dario Pompili

    cs.AI · cs.AR

    The O-RAN ALLIANCE introduces architectural flexibility through disaggregation and AI-driven control. While the RAN Intelligent Controller (RIC) enables centralized optimization, current AI approaches lack reasoning capabilities and explainability. We propose eXplainable RIC (X-RIC), a novel framework integrating agentic AI into O-RAN architectures. X-RIC combines reactive Reinforcement Learning (RL) with strategic agentic reasoning through:...

    arxiv.org/abs/2610.22269 · PDF

  12. 12

    Disagreement-Guided Weak-to-Strong Routing

    Sima Noorani, Shayan Kiyani, George Pappas, Hamed Hassani

    cs.AI

    Small language models are increasingly deployed on edge devices and robotic systems to meet cost and latency constraints, yet their limited capabilities can make reliable deployment challenging. Stronger models can provide assistance when needed, but deciding when to route a weak-model response to a stronger model often relies on signals of response correctness. In open-ended and agentic tasks, however, obtaining such signals can require...

    arxiv.org/abs/2610.22245 · PDF

  13. 13

    Shortcut Sensitivity in Agent-Safety Evaluations

    Hasan Demirkiran, Jens Ernstberger

    cs.AI · cs.CR · cs.LG

    AI agents take actions through tools. Some actions are sensitive, so guards that supervise them must themselves be evaluated reliably. Yet the data used to evaluate a guard can include cues that make labels predictable without measuring the intended safety construct, and those cues can also influence guard decisions. We introduce a three-level shortcut-auditing framework and audit completed LinuxArena trajectories and two TS-Bench-derived...

    arxiv.org/abs/2610.22243 · PDF

  14. 14

    GNN-Guided Local Refinement for Efficient Layout Decomposition with DSA-MP

    Yu Feng, Liyi Zhou, Haoyu Wang, Yiming Xu, Hao Geng, Shaozhi Deng, Ningsheng Xu, Shisheng Xiong

    cs.AI

    Multiple patterning lithography (MPL) has been widely used to improve lithography resolution in advanced nodes, but it faces cost limits as feature sizes shrink. Directed self-assembly (DSA) offers a cost-effective complement, extending the resolution of optical lithography to sub-10 nm and beyond. Integrating DSA into MPL (DSA-MP) further enhances the resolution of contact/via patterning. Nevertheless, growing design complexity exposes the...

    arxiv.org/abs/2610.22214 · PDF

  15. 15

    HTN-Orchestrate: A Hierarchical Task Network Framework for Multi-Agent Orchestration

    Yanchen Deng, Jianguo Huang, Yao Long Teng, Yixuan Li, Handuo Zhang, Ning Wang, Liangcai Su, Xinyu Wang, Bo An

    cs.AI · cs.CL

    Orchestrating multiple agents on a complex task requires reconciling local subtask execution with global task planning. Dynamic dispatch approaches provide execution flexibility but do not explicitly model subtask hierarchies or dependencies, while predefined workflows impose this structure but fix the decomposition before execution begins. Therefore, we introduce HTN-Orchestrate, a two-layer framework that separates these responsibilities...

    arxiv.org/abs/2610.22195 · PDF

  16. 16

    AutoSFT: An Agentic System for Automated Data Research in Supervised Fine-Tuning

    Bingxiang He, Pengqi Sun, Shihao Peng, Shaoxuan Li, Huiyu Xiao, Hongjian Shi, Zixuan Fu, Ziqing Qiao, Yuxin Zuo,...

    cs.AI · cs.CL

    Language model agents increasingly automate post-training, but when given the full training stack they converge on supervised fine-tuning and spend most effort revising training data. Existing systems give the agent one kind of control, selecting rows from a fixed pool or calling a generator, whereas a human data engineer adjusts proportions, prunes, repairs, replaces and synthesizes the data. We design a data action space for these changes...

    arxiv.org/abs/2610.22194 · PDF

  17. 17

    Terminal-RSI: Recursive Self-Improving Terminal Agents via Recursive Self-Rewrite and Policy-Refine

    Junyao Yang, Zongxia Li, Yucheng Shi, Zhongzhi Li, Ruhan Wang, Ruiwen Zhou, Haitao Mi, Shuicheng Yan, Leowei Liang

    cs.AI

    Converting verifier outcomes into useful policy updates and partial progress into complete task resolution is crucial for terminal agents. We present Terminal-RSI, which alternates Recursive Self-Rewrite (RSR) of verified trajectories with critic-inheriting Recursive Policy-Refine (RPR). We formulate Turn-Boundary Skip-Observation GAE (TB-GAE) to retain token-specific baselines while applying decay across interaction turns. Family Dense...

    arxiv.org/abs/2610.22192 · PDF

  18. 18

    Multi-sourced AI Tutor for Non-Native English Speaker Students

    Yiyin Shen, Xinyi Ai, Weibing Wang, Rogers Jeffrey Leo John, Meenakshi Syamkumar

    cs.AI

    Large language models (LLMs) are increasingly being adopted as teaching tools in computing education, particularly for developing AI tutors capable of answering student questions. With the rise in enrollment of international students in computing programs, researchers have begun to explore the difficulties that non-native English speaker (NNES) students face in computing courses due to language barriers. To aid such students, we developed a...

    arxiv.org/abs/2610.22163 · PDF

  19. 19

    Realtime-Bench: Can Frontier Models Solve Real-time Interactive Problems?

    Jingyu Tang, Xu Yang, Zhehao Zhang, Zheng Wang, Qi Cheng, Haozhe Zhao, Manling Li, Wenxuan Li, Xide Xia, Jiaxin Pei

    cs.AI

    Real-time interaction requires agents to perceive evolving environments and act under temporal constraints, yet existing evaluations of real-time streaming models focus on conversational quality and streaming perception, leaving effective action under strict temporal constraints untested. We introduce Realtime-Bench, a suite of carefully curated dynamic tasks where delayed actions directly affect success. Under Wall Clock, where environments...

    arxiv.org/abs/2610.22155 · PDF

  20. 20

    RAG-RST: Retrieved Exemplars as In-Context Discourse Schema for RST Parsing

    Mohammadreza Sediqin, Shlomo Engelson Argamon

    cs.AI

    In Rhetorical Structure Theory (RST), each edge of a discourse tree carries a nuclearity label, marking which span holds the central content, and a relation drawn from a closed inventory of functional categories. Neither label is a surface property: the same two clauses can be Elaboration or Background depending on rhetorical intent, and the boundary between them is fixed by annotation convention rather than by lexical cues. Predicting these...

    arxiv.org/abs/2610.22150 · PDF

  21. 21

    Are You Sure Your LLM Is Sure? Repairing and Auditing Confidence

    Arijit Dey, Bhartendu Kumar, Ramya Hebbalaguppe, Khuushi Maheshwari

    cs.AI

    Confidence scores determine whether an LLM answer is accepted, reviewed, or regenerated, yet scalar recalibration can only correct errors visible through the score itself. We introduce RADAR (Repair, Adopt, and Diagnose After Recalibration), a reliability framework that separates confidence repair from deciding whether to deploy the repair and auditing the errors that remain. Under squared loss, we decompose prediction risk into outcome...

    arxiv.org/abs/2610.22139 · PDF

  22. 22

    Believed but Not Said: Measuring Hidden Belief Collapse in Multi-Agent LLM Discussions

    Liangjie Huang, Henry Peng Zou, Wei-Chieh Huang, Wooseong Yang, Yuwei Han, Philip S. Yu

    cs.AI · cs.CL · cs.MA

    LLM systems increasingly rely on groups of agents that discuss, critique, and check one another, assuming that peers will catch individual errors. We test this assumption against a minimal adversary. In a ten-agent factual discussion, one or three adversarial agents repeatedly assert a plausible false answer. After each round we probe each benign agent's private answer through a separate call whose response is hidden from its peers, and we...

    arxiv.org/abs/2610.22138 · PDF

  23. 23

    CART: Compatible Routing for Multimodal Continual Instruction Tuning

    Ling-Xiang Zhang, Tao Hu, Da-Wei Zhou

    cs.AI

    Multimodal Large Language Models (MLLMs) achieve strong performance through instruction tuning, yet real-world applications often require them to continually acquire new capabilities from sequential tasks, i.e., Multimodal Continual Instruction Tuning (MCIT). In typical MLLMs, a projector is employed to align visual representations with the language model. However, existing methods primarily protect the language model from forgetting while...

    arxiv.org/abs/2610.22137 · PDF

  24. 24

    DropKV: Decoupling Residual-Output Perturbation for Near-Optimal KV-Cache Eviction

    Aozhong Zhang, Selcuk Gurses, Yanxia Deng, Naigang Wang, Chi-Chun Liu, Davis Wertheimer, Derrick Liu, Xin Li, Zi...

    cs.AI

    Key-value (KV) cache management has become a critical bottleneck for scaling Large Language Models (LLMs) to long-context scenarios due to linear memory growth. While recent eviction methods have moved beyond simple attention-score heuristics to incorporate value representations, they typically face a fundamental trade-off: the underlying optimization for minimizing output perturbation is an NP-hard combinatorial problem, leading to...

    arxiv.org/abs/2610.22130 · PDF

  25. 25

    PROBE: Learning to Audit Policy Compliance in Tool-Using LLM Agents

    Kshitij Mishra, Abhijith Sharma, Nils Lukas, Salem Lahlou

    cs.AI

    Modern agentic systems extend language models with external tools, enabling them to complete complex tasks through multi-turn interactions. As these agents are deployed in consequential settings, they are often constrained by policies governing pricing, privacy, confirmation, safety, and other requirements. However, existing evaluation methods leave a critical gap: capability benchmarks measure what agents can accomplish, while content-safety...

    arxiv.org/abs/2610.22124 · PDF

  26. 26

    Learning from Unverified Peers via Shared Residual Distillation

    Feiyang Chen, Runze Ma, Zichen Wang, Jiong Lou, Zihan Wang, Shuangjia Zheng, Jie Li

    cs.AI

    A language model can generate several solutions to the same problem, some leading to correct answers and others to incorrect ones. Many post-training methods turn these trajectories into learning signals using supervision from reference answers, verifiable feedback, or external teachers. Without such supervision, learning from the model's own generations becomes less straightforward, as their correctness cannot be reliably assessed. An...

    arxiv.org/abs/2610.22080 · PDF

  27. 27

    When Reasoning Is Not Enough: Proactive Inquiry for Cross-Modal Safety in Multimodal Large Language Models

    Wei Cai, Jian Zhao, Yuchen Yuan, Xuelong Li

    cs.AI

    Multimodal large language models (MLLMs) can act unsafely when individually benign images and texts are interpreted jointly. Beyond implicit reasoning risk, where reasoning drifts toward harm, we identify intent-level risk, where cross-modal ambiguity silently commits the model to a hazardous premise that path-level optimization alone cannot correct. We propose SAFE-PRO, which uses a Risk-Conditioned Uncertainty (RCU) signal, combining...

    arxiv.org/abs/2610.22068 · PDF

  28. 28

    Prism-1: Position- and Rival-Independent Scoring for System One Models

    Yao Long Teng, Jianguo Huang, Yanchen Deng, Jiayi Cai, Bo An

    cs.AI

    Given a state and a set of options, such as intents, tools or interface elements, System One decision models return a distribution over the options in one forward pass, without generated reasoning. Web systems increasingly build them by reading decisions from language models. Two axioms of rational choice are natural requirements for such decisions, because in web systems the listing and the option set are determined by other components and...

    arxiv.org/abs/2610.22066 · PDF

  29. 29

    FAOPs: Normal Forms and an Algebra for AI Agent Workflows

    Ali Mohammadi Shanghooshabad

    cs.AI · cs.DB · cs.PL

    Organizations increasingly encode operational procedures as multi-step agent skills, here called Agent Operating Procedures (AOPs), that must be traceable, auditable, reproducible, access-controlled, and bounded in time and cost, requirements that regulated sectors make mandatory. Yet AOPs are typically kept as coarse-grained files, in which steps bundle actions, copied procedures drift apart, any step may reach any tool, and refactoring goes...

    arxiv.org/abs/2610.22061 · PDF

  30. 30

    Progress Reward Model for Reinforcement Learning via Large Language Models

    Xiuhui Zhang, Ning Gao, Xingyu Jiang, Yihui Chen, Yuheng Pan, Mohan Zhang, Yue Deng

    cs.AI

    Traditional reinforcement learning (RL) algorithms face significant limitations in handling long-term tasks with sparse rewards. Recent advancements have leveraged large language models (LLMs) to enhance RL by utilizing their world knowledge for task planning and reward generation. However, planning-based approaches often depend on pre-defined skill libraries and fail to optimize low-level control policies, while reward-based methods require...

    arxiv.org/abs/2610.22056 · PDF

  31. 31

    GraphACM: Graph-based Action Consistency Measurement for Uncertainty Quantification in LLM Agents

    Yaxin Ma, Wenbo Zhang, Satya Krishna Pothapragada, Alina Zare

    cs.AI

    Behavioral consistency across repeated executions provides an informative signal for assessing the reliability of large language model (LLM) agents. However, existing approaches typically summarize consistency with a single task-level score, without identifying where behavioral differences occur or whether actions are compared under equivalent decision contexts. We propose a graph-based framework for fine-grained behavioral consistency...

    arxiv.org/abs/2610.22031 · PDF

  32. 32

    CARA: Causally Guided Retrieval Agent for Question Answering

    Weicheng Liu, Xiangcheng Wu, Depeng Xu, Haopeng Zhang, Xi Niu

    cs.AI · cs.IR

    Retrieval-augmented generation over text-rich knowledge graphs requires identifying evidence that contributes to answering a question, rather than merely matching its textual content. Existing agentic retrievers adaptively balance global search and multi-hop exploration but primarily rely on textual similarity to guide retrieval. We propose CARA, a training-free, CAusally guided Retrieval Agent that explicitly evaluates the causal relevance...

    arxiv.org/abs/2610.22017 · PDF

  33. 33

    OntoAct: Hypothesis-Based Neuro-Symbolic Agents through Ontological Abstraction

    Donghyeok Lee, Bong Gyun Kang, Junyong Ahn, Sungroh Yoon

    cs.AI · cs.LG

    Learning from experience requires recovering reusable knowledge and determining when it applies to a new situation. We introduce OntoAct, a hypothesis-based neuro-symbolic framework built on ontological abstraction. OntoAct turns individual experiences into axioms over entity types, relations, and conditions, making their underlying patterns reusable across situations. A large language model (LLM) uses this world knowledge to propose...

    arxiv.org/abs/2610.21983 · PDF

  34. 34

    CoWorld Reasoner: Agentic Multimodal Reasoning with the Evolving Video World Model-Driven Harness

    Dongjie Cheng, Ruifeng Yuan, Ran Wei, Ning Song, Yongqi Li, Wenjie Li

    cs.AI

    As Vision-Language Models (VLMs) advance, they are increasingly expected to perform spatiotemporal reasoning over motion, actions, and view changes beyond what the input images show. Video World Models (VWMs) provide a natural unified engine for visual imagination by generating task-relevant visual content as additional evidence. However, existing approaches typically invoke a VWM through fixed pipelines centered on a predefined form of...

    arxiv.org/abs/2610.21910 · PDF

  35. 35

    RoleQ: Role-Aligned Scales for Converting Language Models to Ternary

    Johannes Knittel, Alec Helbling, Hanspeter Pfister

    cs.AI · cs.CL

    Weight quantization below four bits shrinks memory and bandwidth at a perplexity that looks acceptable, but often at the cost of factual knowledge and task accuracy. Ternary weights have mostly been obtained by training from scratch or long fine-tuning. We introduce RoleQ, a quantization framework that places scales by the functional roles of Transformer units rather than by the layout of the weight matrix, as conventional formats do. Vectors...

    arxiv.org/abs/2610.21808 · PDF

  36. 36

    CIM-Embed: Learning Instruction-Conditioned Multimodal Embeddings for Clustering

    Yinghua Xu, Pengyang Xie, Yanxiang Zhou, Chuanzhi Tong, Xiao Xiao, Chaoqun Feng, Xiaoqing Wang

    cs.AI · cs.LG

    Multimodal objects can require different partitions under different instructions, as when products in an e-commerce catalog are grouped by color, material, or both. Retrieval-oriented embeddings may not recover these partitions, while generative clustering often relies on intermediate descriptions or iterative decisions. This paper presents CIM-Embed, a framework for learning instruction-conditioned multimodal embeddings for clustering....

    arxiv.org/abs/2610.21805 · PDF

  37. 37

    Co2In: a Bi-level Memory Incremental Learning Framework with Knowledge Encoding, Consolidation, and Integration

    Wenju Sun, Qingyong Li, Boyang Li, Wen Wang, Siyu Zhang, Yi Wei, Yangliao Geng

    cs.AI

    Incremental learning (IL) aims to continually acquire new knowledge (plasticity) while retaining previously learned information (stability). However, striking a balance between plasticity and stability remains a significant challenge for intelligent systems. The human brain achieves exceptional balance, owing to various memory units that collaboratively encode and store information. Inspired by human memory mechanisms, this paper introduces...

    arxiv.org/abs/2610.21804 · PDF

  38. 38

    RSIFlow: Adaptive Self-Improvement Workflows with an Evolving Meta-Agent

    Huan Zhang, Mingju Chen, Dongxu Zhou, Can Lv, Sen Cui, Faguo Wu, Shiji Zhou

    cs.AI

    Large language model (LLM) agents have demonstrated strong capabilities in complex tasks, including code generation, tool use, and multi-hop question answering. Existing self-improvement methods enhance agent performance through model optimization, harness evolution, and executable artifact refinement, while recursive self-improvement (RSI) further extends optimization to the improvement mechanisms themselves. However, adapting...

    arxiv.org/abs/2610.21791 · PDF

  39. 39

    From Research Goals to Scientific Closure: Evaluating Agents Through Evidence-Supported Physics Research

    Yijia Mao, Song Dai, Tao Chen, Xiangyu Lu, Linqian Zeng, Chenxi Hu, Sitong Zhou, Dongyang He, Xiaodan Mao, Linxuan...

    cs.AI

    Can AI agents organize scientific investigations that lead to well-supported conclusions? To evaluate this capability, we introduce PhysClosureBench, a benchmark of 50 expert-verified, goal-specified research tasks spanning strong-field, attosecond, and neighboring physics areas. Each task specifies the research objective and essential scientific constraints, while leaving the agent to organize the investigation and determine how to generate...

    arxiv.org/abs/2610.21767 · PDF

  40. 40

    DeciCAD: Harness-Grounded Learning of Action Policies for CAD Agents

    Yufeng Peng, Yixin Zhang, Ziyi Liu, Zhikun Xu, Jiarong Yang, Sijie Wang, Wenjie Feng, Qiyu Kang

    cs.AI · cs.CL · cs.LG · cs.RO

    Large Language Model agents operate through closed-loop action policies whose decision space is defined by an agent harness. We investigate how this operating space shapes agent behavior and policy learning in computer-aided design (CAD), a professional domain characterized by persistent state, geometric constraints, and long-horizon procedures. We instantiate this perspective in DeciCAD, a domain-native harness that exposes semantic CAD...

    arxiv.org/abs/2610.21755 · PDF

  41. 41

    In-Context Preference-Conditioned Pareto Set Learning from Limited Offline Data

    Yiming Yao, Liang Zhao, Qingfu Zhang

    cs.AI

    Multi-objective optimization involves diverse preferences over conflicting objectives. Rather than returning a finite set of trade-off solutions, preference-conditioned Pareto set learning seeks to map a specified preference directly to a corresponding Pareto optimal solution. Yet in many real-world scenarios, optimization must proceed offline without further objective evaluations, leaving only a limited static dataset that sparsely covers...

    arxiv.org/abs/2610.21736 · PDF

  42. 42

    Data World: Scaling Data Environments from Scratch

    Zhanli Li, Zeyuan Xu, Chaoxu Pang, Zeyuan Liu, Anyang Ji, Te Sun, Kongcheng Zhang, Wenbo Xu, Xubin Yue, Shipeng Luo,...

    cs.AI

    Data agents use code to process heterogeneous files and derive analytical insights. Scaling their training faces costly environment curation, privacy and intellectual-property restrictions, and difficulty verifying open-ended findings without known generative truth. Our core insight is to specify the world behind the files before generating the files themselves. We introduce Data World, a multi-agent engine that synthesizes environments,...

    arxiv.org/abs/2610.21656 · PDF

  43. 43

    The Self-Improvement Problem

    Brian B. Moser, Stanislav Frolov, Tobias C. Nauen, Federico Raue, Andreas Dengel

    cs.AI · cs.CL · cs.LG

    With the rise of large language models, agentic systems are becoming increasingly autonomous, as they write code, call tools, revise prompts, and propose modifications to their own components. This autonomy raises a fundamental question for safe self-improvement: can an automated gate like a reward model or an automated reviewer reliably decide whether a proposed update should be accepted? The difficulty is structural. The gate must be a...

    arxiv.org/abs/2610.21646 · PDF

  44. 44

    AgentPerturb: Evaluating LLM Agent Robustness under Controlled Interaction Perturbations

    Yuanjian Xu, Jinyi Han, Jianing Hao, Yuxuan Sun, Fanyang Lu, Richeng Xuan, Guang Zhang, Zhichao Hu

    cs.AI

    Real-world tasks are often complex and noisy, posing substantial challenges for LLM agents. Yet existing evaluations lack a systematic framework for jointly characterizing potential perturbation targets and types. We use this two-dimensional view to conduct a unified study that quantifies the robustness of current frontier models, traces the mechanisms behind perturbation-induced failures, and draws implications for robustness-oriented...

    arxiv.org/abs/2610.21624 · PDF

  45. 45

    Learning Intrinsic Dynamics in Latent World Models

    Jiamin Shi, Yuanhong Zhang, Zhiyang Liu, Jiawei Sun, Jingwen Fu

    cs.AI

    Learning intrinsic dynamics---the relationship between actions and their consequences---is essential for world models. However, visual variations unrelated to the underlying control process can obscure this relationship in learned representations. We investigate the connection between action predictability, defined as the ability to reconstruct an action from consecutive latent states, and the preservation of intrinsic dynamics. Our central...

    arxiv.org/abs/2610.21612 · PDF

  46. 46

    Evaluating Inquiry Learning in LLM-Based Agents: Hypothesizing, Reasoning, and Exploration

    Marta Knežević, Seyed Parsa Neshaei, Jibril Frej, Vinitra Swamy, Tanja Käser

    cs.AI

    Scientific inquiry involves iterative cycles of exploration, hypothesis generation, and reasoning over observations to uncover relationships in complex systems. Large language models (LLMs) have demonstrated strong capabilities in several components of this process, including inductive reasoning, hypothesis generation, and data analysis. However, prior work typically evaluates these abilities in isolation, leaving open how models integrate...

    arxiv.org/abs/2610.21604 · PDF

  47. 47

    LoopAhead: Speculating Ahead for Faster Decoding in Looped Language Models

    Zixuan Liu, Nan Xue, Rui Bao, Yuhang Xu, Xiaofeng Cao, Jiangchao Yao, Zhiyong Chen

    cs.AI · cs.LG

    Looped Transformers improve parameter efficiency by repeatedly applying the same Transformer layers. However, generating each token still involves sequential computation and repeated access to model weights, leading to high decoding latency. Recent pipelined self-speculative decoding methods reduce this delay by drafting tokens from early-loop predictions and asynchronously advancing tokens at different recurrent depths in the same batch, but...

    arxiv.org/abs/2610.21594 · PDF

  48. 48

    LOADBench: A Large Open Analog Dataset and Benchmark for Machine Learning in IC Design

    M. Leibl, F. Azevedo, S. Uhlich, A. Bonetti, A. Venkitaraman, C. Hsieh, P. Pham, L. Servadei, H. Graeb, R. Martins

    cs.AI

    Machine learning (ML) for analog IC design is constrained by limited open data and the absence of standardized evaluation: results reported in different papers are typically obtained on bespoke, often closed, datasets and with ad-hoc metrics, leaving comparisons across methods unreliable. We address both gaps with a large opensource dataset and a suite of standardized benchmarks. The dataset contains more than 100 operational amplifier...

    arxiv.org/abs/2610.21591 · PDF

  49. 49

    ClaimScope: Label-Blind Witness Verification for Mathematical Claim Auditing

    Mushir Akhtar

    cs.AI · cs.LG

    Mathematical agents are commonly evaluated by answer accuracy or proof acceptance. We introduce ClaimScope, a 72-task benchmark for auditing proposed claims through a verdict, scope, counterexample, and repair. Tasks cover optimizer-state transport, finite-horizon affine recurrences, and exact $2\times2$ linear algebra. We evaluate Qwen3-8B with direct generation and four matched two-pass conditions, each starting from the same draft. A...

    arxiv.org/abs/2610.21581 · PDF

This edition is part of The Daily Abstract — cs.AI archive. Subscribe to receive these in your inbox each morning, automatically translated to Spanish, with reply-to-PDF: arxivdaily.ignorelist.com.

Colophon Set in Georgia, with system sans for interface chrome and a monospaced stack for code and paper identifiers. Sole accent: amber #D99C5E. Built and served on an always-free VM. The masthead is set 14% letterspaced because newspapers do that and it works.