AI news archive.
4,965 articles filed under Agents, newest first, from the outlets listed on the sources page.
- arXiv cs.AIagentsSEABench: Benchmarking Endogenous Misalignment In Self-Evolving Agents
arXiv:2609.35596v2 Announce Type: replace-cross Abstract: Self-evolving LLM agents have gained prominence for their ability to improve after deployment by modifying their harness, including their controller instructions,…
- arXiv cs.AIagentsGPUPhysBench: Benchmarking Coding Agents for Correct and Efficient GPU Physics Simulation
arXiv:2609.35639v1 Announce Type: cross Abstract: Writing fast GPU code for physical simulation is difficult: implementations must preserve numerical accuracy while handling irregular data access, synchronization, and it…
- arXiv cs.AIagentsKV-streams for Efficient Compaction in Agentic Reinforcement Learning
arXiv:2609.35750v3 Announce Type: replace-cross Abstract: Scaling the horizon of agentic LLMs is bottlenecked by the need to fit ever longer context traces in GPU memory. Context compaction has been the most popular mech…
- arXiv cs.AIagentsTokenCast: Forecasting Token Consumption During LLM Agent Execution
arXiv:2609.35760v2 Announce Type: replace-cross Abstract: When a large language model (LLM) agent executes the same task, token consumption can vary by over an order of magnitude across runs. The agent chooses its next s…
- arXiv cs.AIagentsBuilding Intelligent Agents with Neuro-Symbolic Concepts
arXiv:2505.06191v2 Announce Type: replace Abstract: This article presents a concept-centric paradigm for building agents that can learn continually and reason flexibly. The concept-centric agent utilizes a vocabulary of…
- arXiv cs.AIagentsEnabling Regulatory Multi-Agent Collaboration: Architecture, Challenges, and Solutions
arXiv:2509.09215v3 Announce Type: replace Abstract: Large language models (LLMs)-empowered autonomous agents are transforming both digital and physical environments by enabling adaptive, multi-agent collaboration. While…
- arXiv cs.AIagentsOn Memory: A comparison of memory mechanisms in world models
arXiv:2512.06983v2 Announce Type: replace Abstract: World models enable agents to plan within imagined environments by predicting future states conditioned on past observations and actions. However, their ability to plan…
- arXiv cs.AIagentsTSR: Trajectory-Search Rollouts for Multi-Turn RL of LLM Agents
arXiv:2602.11767v4 Announce Type: replace Abstract: Advances in large language models (LLMs) are driving a shift toward using reinforcement learning (RL) to train agents from iterative, multi-turn interactions across tas…
- arXiv cs.AIagentsActionEngine: From Reactive to Programmatic Web Agents via State Machine Memory
arXiv:2602.20502v2 Announce Type: replace Abstract: Many web agents operate through a reactive execution loop: they observe the current interface, reason about the next action, execute it, and repeat. This design incurs…
- arXiv cs.AIagentsMASRubric: Auditing Information Flow in Multi-Agent Systems with Failure-Distilled Pitfall Rubrics
arXiv:2602.23258v3 Announce Type: replace Abstract: While multi-agent systems (MAS) excel at complex reasoning, they are vulnerable to errors that intermediate agents introduce and downstream agents build upon. Auditing…
- arXiv cs.AIagentsAgentic Service Markets Across the Computing Continuum: A Polymatroidal Architecture
arXiv:2603.05614v3 Announce Type: replace Abstract: Autonomous AI agents are becoming economic actors. They compose deadline-bound services across the device-edge-cloud continuum and contend for capacity no single operat…
- arXiv cs.AIagentsCommunication Gain and Delay Cost Under Cross-Timestep Delays in Cooperative Multi-Agent Reinforcement Learning
arXiv:2604.03785v3 Announce Type: replace Abstract: Communication is essential for coordination in \emph{cooperative} multi-agent reinforcement learning under partial observability, yet \emph{cross-timestep} delays cause…
- arXiv cs.AIagentsMMORF: A Multi-agent Framework for Designing Multi-objective Retrosynthesis Planning Systems
arXiv:2604.05075v2 Announce Type: replace Abstract: Multi-objective retrosynthesis planning is a critical chemistry task requiring dynamic balancing of quality, safety, and cost objectives. Language model-based multi-age…
- arXiv cs.AIagentsAuditable Agents
arXiv:2604.05485v4 Announce Type: replace Abstract: LLM agents call tools, query databases, delegate tasks, and trigger external side effects. Once an agent system can act in the world, the question is no longer only whe…
- arXiv cs.AIagentsAgentic Forecasting with Structured Linguistic Beliefs
arXiv:2604.18576v5 Announce Type: replace Abstract: We propose the Bayesian Linguistic Forecaster (BLF), which is a search-based ReAct style agent for (binary) forecasting that augments its context with a linguistic 'bel…
- arXiv cs.AIagentsAlignment has a Fantasia Problem
arXiv:2604.21827v3 Announce Type: replace Abstract: In accomplishing complex tasks, human cognition typically progresses from abstract to concrete (e.g., from brainstorming ideas to writing an essay). With the advent of…
- arXiv cs.AIagentsWhat Happens Inside Agent Memory? Circuit Analysis from Emergence to Diagnosis
arXiv:2605.03354v3 Announce Type: replace Abstract: Agent memory failures are silent: an LLM-based agent can produce a fluent response even when it fails to extract, retain, or retrieve the information needed across sess…
- arXiv cs.AIagentsBALAR : A Bayesian Agentic Loop for Active Reasoning
arXiv:2605.05386v2 Announce Type: replace Abstract: Large language models increasingly operate in interactive settings where solving a task requires multiple rounds of information exchange with a user. However, most curr…
- arXiv cs.AIagentsPrefixGuard: Online Failure Warning and Trace-Grounded Diagnosis for LLM Agents
arXiv:2605.06455v2 Announce Type: replace Abstract: Large language model (LLM) agents execute multi-step tasks, but terminal verdicts arrive too late for intervention. Agent traces mix messages, tool calls, and feedback,…
- arXiv cs.AIagentsARMOR: An Agentic Framework for Reaction Feasibility Prediction via Adaptive Utility-aware Multi-tool Reasoning
arXiv:2605.07103v2 Announce Type: replace Abstract: Reaction feasibility prediction, as a fundamental problem in computational chemistry, has benefited from diverse tools enabled by recent advances in artificial intellig…
- arXiv cs.AIagents$\delta$-mem: Efficient Online Memory for Large Language Models
arXiv:2605.12357v2 Announce Type: replace Abstract: Large language models increasingly need to accumulate and reuse historical information in long-term assistants and agent systems. Simply expanding the context window is…
- arXiv cs.AIagentsContext Pruning for Coding Agents via Multi-Rubric Latent Reasoning
arXiv:2605.15315v2 Announce Type: replace Abstract: LLM-powered coding agents spend the majority of their token budget reading repository files, yet much of the retrieved code is irrelevant to the task at hand. Existing…
- arXiv cs.AIagentsLatent Action Reparameterization for Efficient Agent Inference
arXiv:2605.18597v3 Announce Type: replace Abstract: Large language model (LLM) agents often rely on long sequences of low-level textual actions, resulting in large effective decision horizons and high inference cost. Whi…
- arXiv cs.AIagentsForeSci: Evaluating LLM Agents for Forward-Looking AI Research Judgment
arXiv:2606.00644v3 Announce Type: replace Abstract: AI research often requires decisions before future evidence exists: which bottleneck to attack, which direction to pursue, or where a project should be positioned. We i…
- arXiv cs.AIagentsSkillSmith: Co-Evolving Skills and Tools for Self-Improving Agent Systems
arXiv:2606.01314v2 Announce Type: replace Abstract: LLM agents increasingly store capabilities in external skill and tool libraries. Libraries can grow, but execution must fit a finite context window and improvement cons…
- arXiv cs.AIagentsAgentCL: Toward Rigorous Evaluation of Continual Learning in Language Agents
arXiv:2606.02461v3 Announce Type: replace Abstract: Language agents spend substantial inference time solving individual tasks, yet the experience acquired in one episode is often underutilized in future episodes. Continu…
- arXiv cs.AIagentsPRISM: Recovering Instruction Sets from Language Model Activations
arXiv:2606.09563v2 Announce Type: replace Abstract: As LLMs are deployed as agents, reliable monitoring requires knowing not only what they output, but which instructions are steering their behavior. This is difficult wh…
- arXiv cs.AIagentsMind the Perspective: Let's Reason Recursively for Theory of Mind
arXiv:2606.11724v2 Announce Type: replace Abstract: Theory of Mind (ToM) reasoning requires inferring agents' beliefs from partial and asymmetric observations, which remains an open challenge for LLMs. Existing prompting…
- arXiv cs.AIagentsAutonomous Event-Driven Multi-Agent Orchestration for Enterprise AI at Scale
arXiv:2606.20058v2 Announce Type: replace Abstract: Enterprise agent language models must coordinate specialist agents over continuous business event streams, yet most multi-agent evaluations assume discrete request-resp…
- arXiv cs.AIagentsDesign and Implementation of Agentic Orchestrations and Orchestration of Agents
arXiv:2606.31518v2 Announce Type: replace Abstract: Agentic Business Process Management has gained momentum recently. The prospect is that the autonomy of AI agents, i.e., predominantly LLM-based agents, can be balanced…
- arXiv cs.AIagentsOPINE-World: Programmatic World Modeling with Ontology-error-Prioritized Interactive Exploration for ARC-AGI-3
arXiv:2607.01531v3 Announce Type: replace Abstract: Learning a useful world model from minimal interaction is central to building agents that adapt to unfamiliar tasks. Programmatic world modeling approaches such as Worl…
- arXiv cs.AIagentsTurnOPD: Making On-Policy Distillation Turn-Aware for Efficient Long-Horizon Agent Training
arXiv:2607.05804v2 Announce Type: replace Abstract: On-policy distillation (OPD) trains a student policy by matching a stronger teacher on the student's own trajectories, offering a promising framework for language agent…
- arXiv cs.AIagentsTopoExplore: Homology as an Exploration Signal
arXiv:2607.09971v2 Announce Type: replace Abstract: Exploration signals in reinforcement learning are currently computed from what an agent has seen: visitation counts, density estimates, or a model's prediction error at…
- arXiv cs.AIagentsAgents Don't Just Agree, They Remember: Benchmarking Persistent Sycophancy in Self-Improving Personal Agents
arXiv:2607.10526v5 Announce Type: replace Abstract: Self-improving personal agents write profiles, memories, and reusable skills that carry over from one chat to the next. Prior works ask whether a model agrees with the…
- arXiv cs.AIagentsSeekJudge: A Practical Reward Framework for Reinforcement Learning in Computer-Use Agents
arXiv:2607.23263v2 Announce Type: replace Abstract: Deciding whether a trajectory actually fulfills its instruction determines how we measure computer-use agents on long-horizon graphical-user-interface tasks and how we…
- arXiv cs.AIagentsClinLens: Towards Long-Horizon LLM Agents for Longitudinal Multimodal Clinical Data Science
arXiv:2607.26155v2 Announce Type: replace Abstract: Automating clinical data science requires agents to translate patient-centric objectives into executable workflows grounded in longitudinal multimodal evidence. Existin…
- arXiv cs.AIagentsNeSyFS: A Neuro-symbolic Fast-Slow Thinking Framework for LLM Agent under Partial Observability
arXiv:2607.28942v3 Announce Type: replace Abstract: Recently Large Language Models (LLMs) have been increasingly deployed as autonomous agents in applications such as self-reflection, retrieval-augmented generation, and…
- arXiv cs.AIagentsAgentStream: How Well Do Self-Evolving LLM Agents Perform Under Streaming Tasks?
arXiv:2608.00155v2 Announce Type: replace Abstract: Large language model (LLM) agents can self-evolve by continually improving from their own accumulated experience. However, existing studies predominantly adopt independ…
- arXiv cs.AIagentsAssuming You Knew: Fixing an Epistemic Semantics for Flow Policies Using Agentic AI
arXiv:2608.00882v2 Announce Type: replace Abstract: Many high-level security requirements are about the allowed flow of information in programs and are difficult to make precise because they involve selective downgrading…
- arXiv cs.AIagentsModel Discovery Agent: LLM-assisted Bayesian experiment design for data-efficient discovery of mechanistic world models
arXiv:2608.09696v5 Announce Type: replace Abstract: A primary goal of science is to learn mechanistic world models from limited experimental data, both to explain observations and to predict novel interventions. We intro…
- arXiv cs.AIagentsCausal Behavioral Evaluation of AI Agents at Scale via Automated Behavioral Science
arXiv:2608.10030v3 Announce Type: replace Abstract: As AI agents are increasingly deployed in complex and new environments, knowing the conditions that influence their behavior becomes an indispensable step for their rel…
- arXiv cs.AIagentsSpatial Memory Agent: Experience-Grounded Procedural Memory for Spatial Intelligence
arXiv:2608.12743v3 Announce Type: replace Abstract: Spatial intelligence is becoming a foundation for embodied agents, robotic planning, and multimodal assistants. To improve the spatial reasoning ability of VLMs, existi…
- arXiv cs.AIagentsAeroCopilotBench: Safety-Gated Evaluation of LLM Agents on Aircraft Emergency Procedures in an Executable Cockpit
arXiv:2608.16349v2 Announce Type: replace Abstract: Aviation knowledge question answering cannot directly assess the operational effectiveness and safety compliance of large language models throughout aircraft emergency…
- arXiv cs.AIagentsWhat is Missing from AI Post-Training AI: An Empirical Analysis
arXiv:2608.19072v2 Announce Type: replace Abstract: Large language model (LLM) agents can now post-train an LLM end-to-end, raising the prospect of recursive self-improvement (RSI). Yet this progress is measured by aggre…
- arXiv cs.AIagentsOptimal Skill Selection for LLM Agents with Provable Bicriteria Guarantees
arXiv:2608.19993v2 Announce Type: replace Abstract: Loading reusable skill documents into a bounded context window has become a primary way large language model (LLM) agents acquire task-specific capabilities, which make…
- arXiv cs.AIagentsDon't Solve, Just Compare: Tiny Advisors for Runtime Intervention in LLM Agents
arXiv:2608.21027v2 Announce Type: replace Abstract: LLM agents are emerging as an important paradigm for real-world tasks that require reasoning, tool use, and sequential decision-making. As these agents operate over lon…
- arXiv cs.AIagentsLitReview Arena: Evaluating Literature Review Agents with Battle-Style Peer Review Platform
arXiv:2608.21374v2 Announce Type: replace Abstract: Literature reviews are essential to scientific progress, but rigorously evaluating automatically generated reviews remains difficult because many aspects of research ut…
- arXiv cs.AIagentsTraining Needs Trustworthy Worlds: Verified Synthetic Web Environments for Agent Learning
arXiv:2608.21898v2 Announce Type: replace Abstract: Web agents promise to automate complex digital workflows, but their training remains limited by synthetic environments that look plausible while hiding broken links, in…
- arXiv cs.AIagentsLocalLSTC: A Long Short-Term Control Architecture for Locally Deployed GUI Agents
arXiv:2608.25777v2 Announce Type: replace Abstract: Modern GUI-agent frameworks achieve strong desktop task performance with frontier API models, yet persistent control information often remains implicit in growing inter…
- arXiv cs.AIagentsA Systematic Survey of Agentic Skills: Architecture, Lifecycle, and Security
arXiv:2608.29596v2 Announce Type: replace Abstract: Autonomous large language model (LLM) agents increasingly face reliability, context consumption, and execution stability bottlenecks when deployed on complex, long-hori…