Agent技能靠程序性锚定而非知识注入？这篇论文用8135条记录验证，维护技能库的可以看看。
一篇论文分析了Agent技能的实际作用机制，基于8135条标准化试验记录发现，程序性锚定解释了65.7%的技能生效案例，显性知识注入只占4.5%。当技能池规模从5个扩大到100个，实际使用精度从29.6%下降到3.3%。在匹配对比中，技能方法比Workflow Memory高出6.06分，但在脆弱假设、不兼容上下文或适应不足时会失效。该论文发布于arxiv.org/abs/2608.14036。
Interesting paper demystifying agent skills. If you maintain skills for your agent, this one is wor...
Interesting paper demystifying agent skills. If you maintain skills for your agent, this one is worth your time. (bookmark it) Skills are usually assumed to inject knowledge the model lacks. However, this paper finds something interesting. Across 8,135 normalized trial records, procedural anchoring accounts for 65.7% of cases where a skill helps, and explicit knowledge injection accounts for 4.5%. Skills stabilize execution rather than supply facts. As the pool grows from 5 to 100 skills, actual-use precision falls from 29.6% to 3.3%. Skills still beat Workflow Memory by 6.06 points in matched comparisons, and they break under brittle assumptions, incompatible contexts, or insufficient adaptation. Paper: arxiv.org/abs/2608.14036 Track more trending AI papers in our academy: academy.dair.ai 💬 1 🔄 2 ❤️ 18 👀 3439 📊 7 ⚡