腾讯 SkillAdam：让 AI 自动改进智能体技能指令，准确率 28.3% 且 token 消耗降至约三分之一
RT by @rohanpaul_ai: New Tencent paper letting AI auto-improve your agent's instructions works better and costs far less if it remembers past fixes and avoids big, risky rewrites. Agent skills are instruction files that teach an agent how to do a job. Tools that auto-rewrite them often go in circles, burning tokens as new edits undo fixes that already worked. SkillAdam teaches the rewriting AI 2 habits. It keeps a log of what's been fixed, and it makes smaller changes when results are mixed. On long shopping and travel planning tasks, it scored 28.3% average accuracy versus 21.7% for SkillOpt, the best earlier method. It also used about a third as many tokens. If you auto-tune your agent's instructions, give the process a memory of past fixes and a brake on big edits. – arxiv. org/abs/2609.08944 Title: "SkillAdam: Stable and Efficient Skill Evolution for Agents"
腾讯论文提出 SkillAdam，让 AI 自动改进智能体技能指令时记录已修复的问题，并在效果不稳定时只做小幅修改。在长程购物与旅行规划任务上，它平均准确率达 28.3%，高于此前最佳方法 SkillOpt 的 21.7%，token 用量约为后者的三分之一。该方法针对自动改写指令常陷入反复、新改动推翻已有修复的问题。
来源：X：Rohan Paul (@rohanpaul_ai) · x.lingyaoai.com