<?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>激活函数 on 硅基 AGI · 智能体学习与测评</title><link>https://guijiagi.com/tags/%E6%BF%80%E6%B4%BB%E5%87%BD%E6%95%B0/</link><description>Recent content in 激活函数 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, 02 Jul 2026 11:13:00 +0800</lastBuildDate><atom:link href="https://guijiagi.com/tags/%E6%BF%80%E6%B4%BB%E5%87%BD%E6%95%B0/index.xml" rel="self" type="application/rss+xml"/><item><title>Softmax变体2026：从标准到线性</title><link>https://guijiagi.com/posts/softmax-variants-2026/</link><pubDate>Thu, 02 Jul 2026 11:13:00 +0800</pubDate><guid>https://guijiagi.com/posts/softmax-variants-2026/</guid><description>系统梳理Softmax及其变体的数学原理、数值稳定性与在LLM中的应用</description></item><item><title>激活函数综述2026</title><link>https://guijiagi.com/posts/activation-function-survey-2026/</link><pubDate>Thu, 02 Jul 2026 11:03:00 +0800</pubDate><guid>https://guijiagi.com/posts/activation-function-survey-2026/</guid><description>从ReLU到GELU再到SwiGLU，全面梳理激活函数的演进脉络与2026年新趋势</description></item><item><title>激活函数演进：从 ReLU 到 SwiGLU</title><link>https://guijiagi.com/posts/activation-function-evolution/</link><pubDate>Thu, 25 Jun 2026 10:00:00 +0800</pubDate><guid>https://guijiagi.com/posts/activation-function-evolution/</guid><description>梳理 LLM 激活函数的演进历程、数学原理与性能对比</description></item></channel></rss>