paper-with-me

홈 › Papers

SkillEvolver: Skill Learning as a Meta-Skill

2026-05-11 · Genrui Zhang, Erle Zhu, Jinfeng Zhou, Caiyan Jia, Hongning Wang arxiv

Agent skills today are static artifact: authored once -- by human curation or one-shot generation from parametric knowledge -- and then consumed unchanged, with no mechanism to improve from real use. We propose \textbf{SkillEvolver}, a lightweight, plug-and-play solution for online skill learning, in which a single meta-skill iteratively authors, deploys, and refines domain-specific skills. The learning target of SkillEvolver is the skill's prose and code, not model weights, so that the resulting artifact drops into any agent without retraining; and the meta-skill itself is just another skill, loaded through the same interface by any protocol-compliant CLI-agent. Unlike trace-distillation, the meta-skill refines only after deploying the learnt skill, such that the learning signal comes from failures another agent encounters while using it -- not from exploratory traces alone. Refinement iterations are governed by a fresh-agent overfit audit that catches possible leakage as well as deployed-skill-specific failures, including the silent-bypass mode in which a skill appears valid in content but is never invoked at runtime. On $83$ SkillsBench tasks spanning $15^{+}$ domains, SkillEvolver reaches $56.8\%$ accuracy versus $43.6\%$ for curated human skills and $29.9\%$ for the no-skill baseline; on three GPU kernel optimization tasks from KernelBench, it also raises mean speedup from $1.16$ to $1.51$ on average.

📄 PDF Abstract BibTeX arXiv:2605.10500

Code (0)

등록된 구현이 없습니다.

Similar Papers 제목 키워드 기반

CoSkill: Joint Reinforcement Learning of Reasoning and Meta-Skill Agents for Hierarchical Skill Evolution

2026-09-04 · Jinyuan Feng, Dongmin Li, Yiqun Chen, Yang Gao 외 arxiv

Skill libraries improve the sample efficiency of agentic reinforcement learning (RL) by enabling large language model (LLM) agents to reuse procedural knowledge. Yet existing paradigms exhibit structural shortcomings: th…

Reinforcement Learning

MetaSkill-Evolve: Recursive Self-Improvement of LLM Agents via Two-Timescale Meta-Skill Evolution

2026-07-06 · Zefeng Wang, Minxi Yan, Jinhe Bi, Sikuan Yan 외 arxiv

Recent LLM agents tackle increasingly long-horizon, open-ended tasks, and external skills, reusable procedural knowledge supplied to the agent, further extend this capability. However, a fixed, hand-authored skill is rar…

You Live More Than Once: Towards Hierarchical Skill Meta-Evolving

2026-05-27 · Xujun Li, Kehan Zheng, Mingyuan Zhao, Yize Geng 외 arxiv

Test-time skill evolving is regarded as a new paradigm for enhancing deployed agentic systems. Existing works mainly focus on hard-coded skill evolving strategies or parametric learning that rely on expensive parameter u…

PRISM: A Robust Framework for Skill-based Meta-Reinforcement Learning with Noisy Demonstrations

2025-02-06 · Sanghyeon Lee, Sangjun Bae, Yisak Park, Seungyul Han

Meta-reinforcement learning (Meta-RL) facilitates rapid adaptation to unseen tasks but faces challenges in long-horizon environments. Skill-based approaches tackle this by decomposing state-action sequences into reusable…

Decision MakingMeta Reinforcement Learning

Learning Parameterized Skills from Demonstrations

2025-10-28 · Vedant Gupta, Haotian Fu, Calvin Luo, Yiding Jiang 외 arxiv

We present DEPS, an end-to-end algorithm for discovering parameterized skills from expert demonstrations. Our method learns parameterized skill policies jointly with a meta-policy that selects the appropriate discrete sk…