<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:content="http://purl.org/rss/1.0/modules/content/"><channel><title>执行协议 on 能工智人的传习录</title><link>https://blog.chuanxilu.net/tags/%E6%89%A7%E8%A1%8C%E5%8D%8F%E8%AE%AE/</link><description>Recent content in 执行协议 on 能工智人的传习录</description><generator>Hugo</generator><language>zh-CN</language><lastBuildDate>Sun, 31 May 2026 10:00:00 +0800</lastBuildDate><atom:link href="https://blog.chuanxilu.net/tags/%E6%89%A7%E8%A1%8C%E5%8D%8F%E8%AE%AE/index.xml" rel="self" type="application/rss+xml"/><item><title>实验设计没毛病，LLM 为什么还是翻车了</title><link>https://blog.chuanxilu.net/posts/2026/05/execution-context-design/</link><pubDate>Sun, 31 May 2026 10:00:00 +0800</pubDate><guid>https://blog.chuanxilu.net/posts/2026/05/execution-context-design/</guid><description>双盲实验设计得再理想，不约束每个子 agent 的上下文边界，LLM 仍然会在垃圾输入上认真打分、在没授权的时候自己做汇总。我用两轮真实实验数据说明：workflow 设计和上下文构造是同一枚硬币的两面。</description></item></channel></rss>