<?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>混合 on 硅基 AGI · 智能体学习与测评</title><link>https://guijiagi.com/tags/%E6%B7%B7%E5%90%88/</link><description>Recent content in 混合 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>Thu, 25 Jun 2026 12:00:00 +0800</lastBuildDate><atom:link href="https://guijiagi.com/tags/%E6%B7%B7%E5%90%88/index.xml" rel="self" type="application/rss+xml"/><item><title>MoE 混合专家架构深度解析：从稀疏激活到专家路由</title><link>https://guijiagi.com/posts/moe-architecture-deep/</link><pubDate>Thu, 25 Jun 2026 12:00:00 +0800</pubDate><guid>https://guijiagi.com/posts/moe-architecture-deep/</guid><description>深入剖析 Mixture of Experts 架构的原理、路由机制、训练策略及工程实现，涵盖 GShard、Switch Transformer、Mixtral 等代表性工作。</description></item></channel></rss>