Microsoft 论文：CASD 让编码智能体读日志优化提示词，超 GEPA
Microsoft 新论文 CASD，让编码智能体读完整日志自动改提示词，比 GEPA 多提 5.7 分，一次只要 1.6 美元，做 agent 的都该看看。
Microsoft 发表提示词优化论文 CASD，方法是用一个现成的编码智能体读取完整的智能体运行日志，统计错误、找重复失败模式并写成一条提示词，无需环境访问和验证数据。在 ALFWorld、tau2-bench 的 retail 与 telecom、Spreadsheet Bench-Verified 上，单次运行使未优化基线平均提升 16.6 分。对比方法 GEPA 提升 10.9 分、SkillOpt 提升 5.3 分。每条优化后的提示词成本约 1.60 美元，比基于验证集的搜索方法便宜 22 倍以上。
Banger paper from Microsoft on prompt optimization.
(bookmark it)
The claim that a coding agent reading your logs beats GEPA at prompt optimization
The overall finding is that you want to give a coding agent your full set of agent logs and let it write the analysis code, instead of running a search loop over small batches of trajectories.
CASD has an off-the-shelf coding agent compute statistics over the whole trajectory corpus, find recurring failure modes, read representative episodes and write the findings as rules in one prompt. It needs no environment access and no validation data.
Across ALFWorld, tau2-bench retail and telecom, and Spreadsheet Bench-Verified, one pass improves the unoptimized baseline by 16.6 points on average. GEPA improves it by 10.9 and SkillOpt by 5.3.
Each optimized prompt costs about $1.60, more than 22x cheaper than validation-gated search.
Paper: https://t.co/9gp4KNGUTJ
- shao__meng09-24 01:53原文
