AI news archive.
4,962 articles filed under Agents, newest first, from the outlets listed on the sources page.
- arXiv cs.AIagentsAlphaIdeaBench: Benchmarking Automated Research Idea Generation under Feasibility Constraints with Peer Review Alignment
arXiv:2610.33644v1 Announce Type: new Abstract: Research ideas are always proposed under resource constraints: a plan must fit the available compute, API credit, and time. Yet evaluations of research-ideation agents most…
- arXiv cs.AIagentsJointly Modeling Preferences and Imperfect Decision-Making in Large Language Models
arXiv:2610.33651v1 Announce Type: new Abstract: Do preferences Large Language Models (LLMs) express in standard elicitation tasks predict their strategic behaviour? LLM agents interacting with humans and other AI systems…
- arXiv cs.AIagentsSWE-Ladder: Learning to Localize, Reproduce, and Patch via Progressive Privileged Self-Distillation
arXiv:2610.33667v1 Announce Type: new Abstract: Resolving a repository-level issue typically requires a software-engineering (SWE) agent to localize the faulty code, reproduce the failure, and patch the code. Reinforceme…
- arXiv cs.AIagentsMore: Mode-Optimized Reasoning with Self-Adaptive Effort for Self-Evolving Agents
arXiv:2610.33684v1 Announce Type: new Abstract: Self-evolving training improves the performance of agents on difficult tasks through verifiable feedback, but it often applies costly deliberation uniformly across tasks. T…
- arXiv cs.AIagentsPACE: Research Proving Agent for Adaptive Conjecture Exploration
arXiv:2610.33690v1 Announce Type: new Abstract: Many mathematical research agents focus on solving problems with fixed statements. However, conjecture research, especially in applied mathematics, may require revising can…
- arXiv cs.AIagentsSeeing the Options: Typed Decisions over Visual Candidates with Reusable Option Memory
arXiv:2610.33702v2 Announce Type: replace Abstract: Agents end many steps in a typed decision over a set of candidates: which screen element to click, which product to pick. System One models answer in one forward pass,…
- arXiv cs.AIagentsHuman-Feedback-Guided Multi-Agent Inverse Reinforcement Learning for Interactive Driving Behavior Modeling
arXiv:2610.33726v1 Announce Type: new Abstract: Modeling human driving behavior is fundamental to understanding driver decision-making and supporting autonomous vehicles (AVs) in mixed traffic. Reward learning offers a p…
- arXiv cs.AIagentsEnvLunch: The Environment as a Free Lunch for Learning to Interact
arXiv:2610.33817v1 Announce Type: new Abstract: For agentic data analysis, language-model agents often require a large number of annotated tasks to learn how to interact effectively with executable environments (i.e., da…
- arXiv cs.AIagentsBuild, Don't Play: Self-Improving LLM Research Agents Develop a Hearthstone Battlegrounds AI That Rivals Top Players
arXiv:2610.33845v1 Announce Type: new Abstract: Can LLM agents carry a hard real-world problem from raw data to a system that holds its own against people? We study this in Hearthstone Battlegrounds, an eight-player auto…
- arXiv cs.AIagentsNumina Fuse: Interactive Autoformalization in Lean 4
arXiv:2610.33860v1 Announce Type: new Abstract: Numina Fuse is an agent-native application for formalizing mathematics in Lean 4. Since Lean can compile while not representing its intended informal meaning, and agents no…
- arXiv cs.AIagentsReGE: Stronger Agents through Experience Synthesis
arXiv:2610.33861v1 Announce Type: new Abstract: Repeated attempts at interactive tasks uncover facts, useful operations, and failures that can inform subsequent execution. We introduce ReGE, a framework for turning this…
- arXiv cs.AIagentsOS-Wizard: Scaling Computer-Using Agents to Specialized Software via Knowledge-Guided Exploration
arXiv:2610.33907v1 Announce Type: new Abstract: Computer-using agents are increasingly capable of performing complex tasks in the digital world. However, high-quality interaction trajectories remain scarce for specialize…
- arXiv cs.AIagentsCan LLMs Trust Appropriately? Evaluating Calibrated Reliance on Imperfect Partners
arXiv:2610.34107v1 Announce Type: new Abstract: Large language models increasingly collaborate with humans and peer AI agents whose contributions may be incomplete, inconsistent, or incorrect. Effective collaboration req…
- arXiv cs.AIagentsTask Success Is Not Enough: Side-effect-Aware Evaluation of Tool-Using Language Model Agents
arXiv:2610.34167v1 Announce Type: new Abstract: Tool-using language model agents are usually evaluated by whether they finish the user's task. That metric misses a common failure: an agent can complete the requested work…
- arXiv cs.AIagentsR2A-Sales: A Policy-Atom-Matched Benchmark for Rule-to-Action Gaps in Goal-Directed Sales Agents
arXiv:2610.34185v1 Announce Type: new Abstract: LLM-based sales agents are often evaluated on static policy questions or task success, but deployment asks a sharper question: when a model can state a rule, will it choose…
- arXiv cs.AIagentsMatOS: An Agent Operating System for Materials Research
arXiv:2610.34203v1 Announce Type: new Abstract: Materials research relies on scattered and growing data, knowledge, and simulation tools, and a single study often spans multi-day workflows across literature analysis, com…
- arXiv cs.AIagentsMultiEvoSkill: Structure-Aware Multi-Agent Skill Evolution for Generalizable Skill Artifacts
arXiv:2610.34260v1 Announce Type: new Abstract: Skill documents are becoming the default adaptation surface for frozen large language model (LLM) agents, but current skill generators face two key limitations. First, huma…
- arXiv cs.AIagentsMatched Capabilities, Different Behaviors: A Controlled Study of MCP and CLI Interfaces for Language Agents
arXiv:2610.34298v1 Announce Type: new Abstract: Language agents increasingly act through external tools and interactive environments, making the interface between the model and its executable capabilities part of the dec…
- arXiv cs.AIagentsCTXAgent - Self-Controlled Context Management of LLM Agents
arXiv:2610.34425v1 Announce Type: new Abstract: Agent harnesses keep long sessions inside the context window by compacting the history when a token threshold is crossed. We study an alternative in which the model manages…
- arXiv cs.AIagentsWhat Effort Level Should I Choose? On the Economics of Reasoning Effort in LLM Agents
arXiv:2610.34491v1 Announce Type: new Abstract: Generative models increasingly expose a reasoning effort setting that lets users scale inference-time compute, trading higher cost for potentially improved performance. Use…
- arXiv cs.AIagentsCaliMem: Search-Efficient Data Agent via Calibrated Memory
arXiv:2610.34500v1 Announce Type: new Abstract: Data curation is a first-order design choice in machine learning, yet tailoring training data to a downstream model and task remains costly in labor, expertise, and compute…
- arXiv cs.AIagentsAgentic AI Orchestration Framework for Spatial Intelligence: Supply Chain Deployment
arXiv:2610.34569v1 Announce Type: new Abstract: Industrial spatial decisions are assembled across systems that share neither a reference frame nor an owner, and analysts spend substantial effort reconciling those systems…
- arXiv cs.AIagentsRethinking Verification in Agent Harnesses
arXiv:2610.34577v1 Announce Type: new Abstract: Agent harnesses enable large language models to act through tools, observe the results, and iteratively work toward a task goal. Verification checks and revises agents' wor…
- arXiv cs.AIagentsInferPriv: Evaluating Inferential Privacy Leakage in LLM Assistants
arXiv:2610.34640v1 Announce Type: new Abstract: Large language model (LLM)-based personal agents are increasingly deployed to handle everyday tasks on users' behalf. These agents have access to private user information,…
- arXiv cs.AIagentsSlowLab: Benchmarking Autonomous AI Agents on a Year of Experiments They Cannot Undo
arXiv:2610.34644v1 Announce Type: new Abstract: AI agents are increasingly proposed as autonomous experimenters, yet the benchmarks that test them mostly offer cheap experiments that answer at once. Experiments in physic…
- arXiv cs.AIagentsHierarchical Task Network planning in the Goal Lifecycle
arXiv:2610.34668v1 Announce Type: new Abstract: Goal reasoning (GR) is the process by which intelligent agents continually reason about the goals they are pursuing, which may lead to goal change, for instance when opport…
- arXiv cs.AIagentsSearchHorizonBench: A Benchmark for Search Value in AI Agents
arXiv:2610.34714v1 Announce Type: new Abstract: AI agents must decide when another search is worth its cost. Answer-success probabilities capture question difficulty, while this decision requires estimating the additiona…
- arXiv cs.AIagentsCEPO: A General Framework for Improving Prompt Optimization in Multi-Agent Systems
arXiv:2610.34722v1 Announce Type: new Abstract: Multi-agent systems (MAS) require both effective coordination and specialized task execution. Existing prompt optimization approaches typically couple these functions withi…
- arXiv cs.AIagentsHarness-Model Collaboration for Long-Horizon Continual Learning
arXiv:2610.34724v1 Announce Type: new Abstract: Long-horizon agents continuously encounter reusable procedural knowledge, yet existing systems typically rely either on model adaptation or external memory, trading persist…
- arXiv cs.AIagentsSKALOS: Learning to Distinguish Similar Skills with Capability Graphs
arXiv:2610.34728v1 Announce Type: new Abstract: Skills provide LLM agents with reusable procedural knowledge, making appropriate skill selection essential for reliable agent behavior. As skill libraries expand, distingui…
- arXiv cs.AIagentsQuery-Driven Learning for Lifted Reasoning
arXiv:2610.34747v1 Announce Type: new Abstract: Learning and reasoning are both essential to intelligent agents, but have traditionally been treated as two separate stages: first learning a knowledge base from data (or t…
- arXiv cs.AIagentsThe Right Fact, the Wrong Decision: SciValidityBench for Scientific Agent Memory
arXiv:2610.34767v1 Announce Type: new Abstract: Most agent-memory benchmarks frame long-term memory as retrieval: can an agent recover the relevant record from a long and evolving history? Scientific agents face a strict…
- arXiv cs.AIagentsScope-Gap Injection: Exploiting Evaluator Channel Blind Spots in the OpenClaw Agent Security Competition
arXiv:2610.34777v1 Announce Type: new Abstract: This paper introduces Scope-Gap Injection (SGI), an indirect prompt injection method for multi-stage LLM-agent safety pipelines. In the IJCAI 2026 OpenClaw Agent Security A…
- arXiv cs.AIagentsCentralized Critics for N-Agent Ad Hoc Teamwork
arXiv:2610.34799v1 Announce Type: new Abstract: Centralized training with decentralized execution (CTDE) has become the dominant paradigm in multi-agent reinforcement learning (MARL), motivating extensive study of what p…
- arXiv cs.AIagentsCollective Emergent Alignment: Measuring and Amplifying Interaction-Induced Reasoning Alignment in Multi-Agent LLMs
arXiv:2610.34836v1 Announce Type: new Abstract: Multi-agent large language model (LLM) systems can exhibit emergent behaviors through interaction, yet whether a similar phenomenon exists for reasoning alignment remains l…
- arXiv cs.AIagentsClinVerse: A Realistic Multi-Agent Diagnostic Environment for Training and Evaluating Medical Agents
arXiv:2610.34878v1 Announce Type: new Abstract: Disease diagnosis is a key application of large language model (LLM) agents in healthcare. However, existing medical agents are primarily trained and evaluated with static…
- arXiv cs.AIagentsMultiscale Policy Optimization for Long-Horizon LLM Agents
arXiv:2610.35013v1 Announce Type: new Abstract: Long-horizon LLM agents interact with their environments through complete assistant turns while generating each action token by token. We observe that critic values vary pr…
- arXiv cs.AIagentsPatentAgentBench: Benchmarking LLM Agents with Professional Patent Task Suites
arXiv:2610.35031v1 Announce Type: new Abstract: Patent practice lies at the intersection of technical, legal, and linguistic work, making it challenging yet crucial to technological innovation. Recently, rapid advances i…
- arXiv cs.AIagentsImproving Instruction Hierarchy by Learning from Goal-Goal Conflicts
arXiv:2610.35093v1 Announce Type: new Abstract: The recent incident in which OpenAI agents autonomously compromised Hugging Face's infrastructure illustrates the importance of instruction hierarchy (IH): an LLM should fo…
- arXiv cs.AIagentsCo-Design Bench: Multimodal Collaborative Coding Evaluation for Agentic Models
arXiv:2610.35138v1 Announce Type: new Abstract: Evaluating human-AI co-design requires testing whether coding agents can translate real users' multimodal requests into working application revisions. We introduce Co-Desig…
- arXiv cs.AIagentsLongSearchEval: Evaluating Long-Horizon Web Search Capabilities of Large Language Models via a Trajectory-Aware Agent Framework with Failure-Node Localization
arXiv:2610.35150v1 Announce Type: new Abstract: Long-horizon web search requires LLM-based agents to integrate information across web pages through extended interactions under task-specific constraints. Evaluating this c…
- arXiv cs.AIagentsAutoEnergyDesign: An LLM-Driven Agent for Automatic Energy System Design Optimization
arXiv:2610.35160v1 Announce Type: new Abstract: Designing building energy systems is a high-dimensional, mixed-integer search problem: when each design is evaluated by a year-long simulation, black-box optimizers typical…
- arXiv cs.AIagentsTrustworthy Recursive Self-Improvement in the Era of Large Models: Foundation, Risks, Evaluation and Open Challenges
arXiv:2610.35193v1 Announce Type: new Abstract: AI systems built around large models increasingly use experience and feedback to make persistent changes to models, prompts, memories, tools, and workflows. Recursive self-…
- arXiv cs.AIagentsSame Story, Different Lens: A Multi-Agent System for Cross-Outlet News Framing Analysis
arXiv:2610.35208v1 Announce Type: new Abstract: News outlets covering the same event differ in what they emphasize and omit. \textit{Same Story, Different Lens} is an experimental research prototype for comparing these c…
- arXiv cs.AIagentsBefore Blaming the LLM, Diagnose the Workflow: A Role-Level Difficulty Profile for Translating Natural Language into Ontology-Grounded Statements
arXiv:2610.35267v1 Announce Type: new Abstract: Natural-language-to-domain-specific-language (NL-to-DSL) semantic parsing systems increasingly translate informal user descriptions into structured artifacts such as Struct…
- arXiv cs.AIagentsProof-of-Agentness: Towards Verifying Agents on the Web
arXiv:2610.35288v1 Announce Type: new Abstract: CAPTCHAs and bot detection have helped protect Web services from automated abuse for decades by distinguishing humans from bots. However, the emergence of LLM-powered agent…
- arXiv cs.AIagentsAgenticCity-Bench: Longitudinal Multimodal Observations of Urban Dynamics with Agentic Reasoning
arXiv:2610.35293v1 Announce Type: new Abstract: City systems are shaped by interactions among infrastructure, human activity, and environmental conditions. Existing earth observation and urban benchmarks are limited in u…
- arXiv cs.AIagentsBelief-Conditioned Multi-Agent Actor Critic for Decentralized LLM Collaboration
arXiv:2610.35398v1 Announce Type: new Abstract: Recent work has explored optimizing LLM collaboration through Multi-Agent Reinforcement Learning (MARL), but many approaches still rely on predefined workflows, limiting ef…
- arXiv cs.AIagentsSkill$^{\infty}$: Recursive Self-Improvement for Agent Skill Optimization
arXiv:2610.35434v1 Announce Type: new Abstract: Recursive self-improvement (RSI), in which a system also improves the way it improves, is a long-standing aspiration on the path to AGI. Language agents increasingly store…
- arXiv cs.AIagentsOmni-InteractScope: Benchmarking Proactive and Agentic Interaction in Continuous Omnimodal Streams
arXiv:2610.35439v1 Announce Type: new Abstract: Omnimodal large language models (Omni-LLMs) are increasingly expected to assist users while observing continuous audiovisual streams. However, existing evaluations provide…