AIAny
Icon for item

Agent-Editing World Model: Rethinking World Modeling for LLM Agents

Edits an LLM agent's internal task state and action continuations to remove unsupported assumptions and noisy plans, improving downstream decisions instead of reconstructing high-entropy tool outputs. Combines Action Judge, State Revision, and EditAct and shows consistent gains across benchmarks and agent backbones.

Introduction

Most LLM-based agents try to predict environment or tool outputs, but those high-entropy, execution-dependent predictions often add little value when true feedback is available. Worse, agents accumulate "task-state contamination": stale assumptions and unsupported plans that linger in history and distort later decisions. This paper flips the world-modeling goal from reconstructing outputs to editing the agent's internal state and future reasoning traces so that subsequent choices are grounded in verifiable progress.

Key Findings
  • Action-oriented editing beats next-observation simulation: AEWM trains an Action Judge to label decisions as Critical, Exploratory, or Noisy, then applies State Revision to rewrite noisy continuations. So what? Agents exposed to these edits make fewer misinformed commitments and recover from bad plans faster.

  • Integration with execution improves outcomes: EditAct ties edits to real executions rather than only issuing critiques, directly changing the state that subsequent reasoning uses. So what? This produces measurable downstream gains—EditAct raises average benchmark scores by about 3.2–6.7 points across three agent backbones.

  • Better supervision and fine-tuning recipes: The paper introduces AEWM-RFT, a rejection-sampling fine-tuning on verified EditAct trajectories that improves over Self-RFT by ~2.2–2.6 points without needing online AEWM guidance. So what? You can obtain most benefits offline through curated, edited trajectories.

  • Robust Action Judge performance: AEWM attains 70.5% macro-F1 on the introduced Action Judge benchmark, outperforming the strongest baseline by 10.6 points. So what? Reliable classification of decision types is central to effective state edits.

Who it's for and trade-offs

Great fit if you develop or evaluate LLM-based agents that rely on long-horizon planning, tool use, web search, or software engineering workflows and find that hallucinated or stale reasoning history harms performance. The approach is particularly relevant when real execution feedback is intermittently available and you prefer editing agent state over simulating noisy outputs.

Look elsewhere if your setting provides perfect, low-latency environment simulators or you cannot alter an agent's decision history at inference time—AEWM's gains depend on the ability to integrate edits into the agent's subsequent reasoning and, in some variants, on curated fine-tuning data.

Where it fits

AEWM sits between simulation-heavy world models and purely critique-based repair: instead of predicting what tools will return, it judges which decisions matter, edits noisy continuations, and commits those edits into the agent's state so planning proceeds from a corrected history. This reframing is especially useful for mitigating "task-state contamination" in long-horizon, multi-tool interactions.

Information

  • Websitearxiv.org
  • OrganizationsAffiliation: Gaoling School of Artificial Intelligence, Renmin University of China
  • AuthorsShuang Sun, Guoxin Chen, Fanzhe Meng, Jia Deng, Huatong Song, Jinhao Jiang, Wayne Xin Zhao, Hongteng Xu, Ji-Rong Wen
  • Published date2026/09/23

More Items

Alternates a Planner (issues sub-queries) and a Synthesizer (integrates retrieved evidence into a persistent summary) to tackle long-horizon deep-search; introduces Role‑Decoupled Policy Optimization (RDPO) for role-specific RL credit assignment and shows strong results (IterSynth-8B reaches 50.7% on five benchmarks).

A mobile-planning agent paper that develops a closed-loop AI-for-AI lifecycle to generate data, train a planner model, and co-evolve a runtime Harness for multi-app task execution. Demonstrates top performance on MobilePA-Bench with improved tool use, memory, skill coordination and low estimated per-task output cost.

Provides a unified benchmark and a spatially aware omni-modal model for spatial audio-visual reasoning in embodied agents — OmniEchoBench (197 scenes, 2,972 QA pairs; 900 FOA-based navigation tasks) plus OmniEcho, which adds an FOA spatial encoder to a pretrained semantic audio pathway to enable spatial QA and sound-guided navigation.