cs.CLNov 16, 2026

MetaAct: Self-Evolving Executable Action Constraints for Embodied Language Agents

Authors: Jiaxing Zhao, Hanwen Wang, Di Liu, Shaoteng Liu, Jing-Ran Su, Yanheng Li, Qian Zha, Jiaxi Li, 

Organizations: Jilin University · Southern University of Science and Technology · Northwestern Polytechnical University · The Hong Kong Polytechnic University · 

Abstract

Self-evolving language agents have emerged as a promising paradigm for continuously improving agent behavior through experience. However, most existing approaches evolve textual artifacts, which can influence action preferences but cannot reliably enforce action admissibility in embodied environments with discrete and often irreversible actions. We propose MetaAct, which shifts self-evolution from textual guidance to executable action constraints by synthesizing GEAR (Grounded Executable Action Rules) from execution feedback. Each rule is represented as a typed specification, compiled into a deterministic guard, and validated offline through trajectory replay without additional model calls or environment interactions. Across four benchmarks, MetaAct improves task performance; on ScienceWorld and WebShop, violations drop from 70% to 0% and 40% to 1%, with score gains of +10.1 and +8.0. Our code is available at https://github.com/BaiCubed/MetaAct.

Figures & tables

Appendix figures & tables13 assets

Supplementary material from the paper’s appendix.

Appendix

Explore similar work

CardsList
  1. EnactOpt: Diagnosis-Guided Harness Evolution for Embodied Agents

    Oct 10, 2026ZhiYuan Feng, Zhuoshi Pan, Zhaolu Kang +5Agent Harness OptimizationEmbodied AI

  2. EmbodiedEvo: Self-Evolving Coding Agent For Physical Intelligence

    Oct 13, 2026Wenyi Li, Yizhe Chi, Youjie Zheng +8Self-Improving AgentsRobot Skill Learning

  3. EmbodiSkill: Skill-Aware Reflection for Self-Evolving Embodied Agents

    May 11, 2026Ruofei Ju, Xinrui Wang, Xin Ding +12LLM Agent Skill LearningAgent Skill Learning