<?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>偏好对齐 on 硅基 AGI · 智能体学习与测评</title><link>https://guijiagi.com/tags/%E5%81%8F%E5%A5%BD%E5%AF%B9%E9%BD%90/</link><description>Recent content in 偏好对齐 on 硅基 AGI · 智能体学习与测评</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>Tue, 30 Jun 2026 09:40:00 +0800</lastBuildDate><atom:link href="https://guijiagi.com/tags/%E5%81%8F%E5%A5%BD%E5%AF%B9%E9%BD%90/index.xml" rel="self" type="application/rss+xml"/><item><title>DPO偏好对齐训练：数据工程与超参调优</title><link>https://guijiagi.com/posts/dpo-preference-alignment-guide/</link><pubDate>Tue, 30 Jun 2026 09:40:00 +0800</pubDate><guid>https://guijiagi.com/posts/dpo-preference-alignment-guide/</guid><description>DPO已成为偏好对齐的主流方案，本文详解DPO的数据构建、训练配置与超参调优全流程</description></item><item><title>DPO 训练实践：偏好对齐的数据工程</title><link>https://guijiagi.com/posts/dpo-training-practice-preference-alignment-data-engineering/</link><pubDate>Sun, 28 Jun 2026 10:00:00 +0800</pubDate><guid>https://guijiagi.com/posts/dpo-training-practice-preference-alignment-data-engineering/</guid><description>深入解析 DPO（Direct Preference Optimization）训练流程，涵盖偏好数据构建、训练配置和效果评估</description></item></channel></rss>