<?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>自我批评 on AI 实战派 · 从技术到赚钱</title><link>https://guijiagi.com/tags/%E8%87%AA%E6%88%91%E6%89%B9%E8%AF%84/</link><description>Recent content in 自我批评 on AI 实战派 · 从技术到赚钱</description><generator>Hugo</generator><language>zh-cn</language><copyright>本站内容采用 CC BY-NC-SA 4.0 国际许可协议授权</copyright><lastBuildDate>Thu, 08 Oct 2026 10:00:00 +0800</lastBuildDate><atom:link href="https://guijiagi.com/tags/%E8%87%AA%E6%88%91%E6%89%B9%E8%AF%84/index.xml" rel="self" type="application/rss+xml"/><item><title>撤回三份、修订十四份：可验证奖励正在训练 AI 的「自我批评」</title><link>https://guijiagi.com/posts/2026-10-08-rlvr-truthfulness-virtue/</link><pubDate>Thu, 08 Oct 2026 10:00:00 +0800</pubDate><guid>https://guijiagi.com/posts/2026-10-08-rlvr-truthfulness-virtue/</guid><description>OpenAI 发布数学成果当天撤回 3 份手稿、修订 14 份。本文讨论以可验证奖励塑形模型，如何逼出一种机器层面的求真德性。</description></item></channel></rss>