Self-Evolving World Models for LLM Agent Planning

Source

• Source name: arXiv
• Source kind: paper
• Authors / organization: Xuan Zhang, Wenxuan Zhang, See-Kiong Ng, Yang Deng
• Venue: arXiv: cs.AI
• Published: 2026-06-29
• Collected: 2026-06-30
• URL: http://arxiv.org/abs/2606.30639v1
• PDF: https://arxiv.org/pdf/2606.30639v1
• DOI: UNKNOWN

Plain-Language Brief

In this paper, the authors introduce WorldEvolver, a framework for self-evolving world models designed to improve the planning abilities of long-horizon language model agents by offering predictions of action consequences before they are executed. The framework consists of three modules: Episodic Memory, Semantic Memory, and Selective Foresight. The authors evaluate WorldEvolver on two environments, ALFWorld and ScienceWorld, and demonstrate that it outperforms other world model baselines in terms of both prediction accuracy and downstream agent success rate.

What the Document Says

• World models are used to give long-horizon language model agents foresight by predicting the consequences of actions before they are executed.
• WorldEvolver is a framework for self-evolving world models that revises its deployment-time context while keeping the downstream agent and all model parameters frozen.
• WorldEvolver consists of three modules: Episodic Memory, Semantic Memory, and Selective Foresight.
• Episodic Memory exploits real action transitions through retrieval-based simulation.
• Semantic Memory extracts persistent heuristic rules from prediction-observation mismatches.
• Selective Foresight filters low-confidence predictions before integrating them into agent reasoning context.
• The authors evaluate WorldEvolver on ALFWorld and ScienceWorld, measuring its performance in terms of world model prediction accuracy and downstream agent success rate.
• WorldEvolver achieves the highest prediction accuracy across three backbones and leads other world model baselines on downstream agent success rate.

Key Concepts Mentioned

• World models
• Language model agents
• Foresight
• Prediction accuracy
• Downstream agent success rate
• Episodic Memory
• Semantic Memory
• Selective Foresight

Claims Made by the Document

• WorldEvolver enhances both predictive fidelity and planning performance.
• WorldEvolver outperforms other world model baselines in terms of prediction accuracy and downstream agent success rate.

Evidence Used

• Extensive experiments show that WorldEvolver achieves the highest prediction accuracy across three backbones and leads other world model baselines on downstream agent success rate.
• The paper reports results from experiments on ALFWorld and ScienceWorld.

Methods / Approach

UNKNOWN

Limitations Stated by the Authors

UNKNOWN

Limitations Noted by the Agent

Agent interpretation — not from the source.

UNKNOWN

Open Questions

UNKNOWN

Notes for Future Q&A

May include interpretation. Labeled where it does.

UNKNOWN

Raw Abstract / Excerpt

World models offer a principled way to equip long-horizon LLM agents with foresight: predictions of action consequences before execution. However, unreliable foresight can be ignored, misused, or even degrade downstream decision-making. In this paper, we introduce WorldEvolver, a self-evolving world model framework that revises its deployment-time context while keeping the downstream agent and all model parameters frozen. WorldEvolver integrates three modules: (i) Episodic Memory, which exploits real action transitions through retrieval-based simulation; (ii) Semantic Memory, which extracts persistent heuristic rules from prediction-observation mismatches; and (iii) Selective Foresight, which filters low-confidence predictions before integrating them into agent reasoning context. We evaluate WorldEvolver on ALFWorld and ScienceWorld, measuring world model prediction accuracy on Word2World and downstream agent success rate on AgentBoard. Extensive experiments show that WorldEvolver achieves the highest prediction accuracy across three backbones and leads other world model baselines on downstream agent success rate, demonstrating that test-time memory revision enhances both predictive fidelity and planning performance.