# Google 新论文：WikiSkill 将智能体经验编译为持久知识库

- 来源：DAIR.AI (@dair_ai)
- 发布时间：2026-08-28 21:06
- AIHOT 分数：47
- AIHOT 链接：https://aihot.virxact.com/items/cmtd06o7z09xqroq5yhhdcfzb
- 原文链接：https://x.com/dair_ai/status/2093324233158045788

## AI 摘要

Google 发布论文 WikiSkill，将智能体技能进化系统拆分为原始执行轨迹、持久知识 wiki 与可执行技能三部分，经验先沉淀进 wiki，后续技能更新基于该知识库。消融实验显示 wiki 是主要增益来源：小模型配进化技能可超越规模大得多的无技能模型，且一个模型进化出的技能可跨模型家族迁移，有时甚至优于自进化技能。论文地址：arxiv.org/abs/2608.27454。

## 正文

Banger paper from Google.

If you maintain a skill library for your agents, you might want to check this out.

(bookmark it)

This work separates three things that skill-evolution systems usually collapse into one. Raw execution traces, a persistent wiki of accumulated knowledge, and the executable skills themselves.

Experience gets consolidated into the wiki, and every later skill update builds on that wiki instead of on a scattered optimization history.

Ablations confirm the wiki is what carries a lot of the gain. Two results stand out in particular. Smaller models with evolved skills beat substantially larger models without them. And skills evolved by one model transfer across families, where skills evolved elsewhere sometimes beat self-evolved ones.

Paper: https://arxiv.org/abs/2608.27454

Chat with Paper: https://academy.dair.ai/papers/wikiskill-compiles-agent-experience-into-a-persistent-wiki-2608.27454