[论文] WikiSkill: Compiling Agent Experience into Persistent Knowledge for Sk...

论文概要

研究领域: NLP
作者: Liyan Tang, Cyrus Rashtchian, Chun-Sung Ferng, Andrew Tomkins, Da-Cheng Juan, Tu Vu
发布时间: 2026-08-27
arXiv: 2608.27454

中文摘要

智能体技能将专业知识和工作流程打包成可重用的资源，扩展AI智能体的能力。近期研究从智能体经验中自动发现此类技能，使智能体能够通过交互逐步适应。然而，指导技能开发的洞察通常分散在优化历史记录中，限制了跨迭代的系统复用。我们引入了WikiSkill，一个将智能体技能与持久知识库（wiki）共同演化的框架。高层来看，WikiSkill将原始执行经验、累积知识和可执行技能分离，同时将经验持续整合到wiki中，后续技能更新可以在此基础上构建。在多个基准测试和模型上，WikiSkill始终优于最先进的技能演化方法，并在大多数模型-基准设置中优于无技能基线。我们发现技能演化与模型扩展互补：较大模型通常从演化技能中获益更多，而较小模型配合技能可以大幅超越无技能的更大模型。

原文摘要

Agent skills package specialized knowledge and workflows into reusable resources that extend AI agent capabilities. Recent work automatically discovers such skills from agent experience, which enables agents to progressively adapt through interaction. However, the insights that guide skill development typically remain scattered across optimization histories, limiting their systematic reuse across iterations. We introduce WikiSkill, a framework that co-evolves agent skills with a persistent knowledge base (wiki). At a high level, WikiSkill separates raw execution experience, accumulated knowledge, and executable skills, while continuously consolidating experience into the wiki, which subsequent skill updates can build on. Across diverse benchmarks and models, WikiSkill consistently outperfo...

自动采集于 2026-08-30

#论文 #arXiv #NLP #小凯