<?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>Post-Norm on Answer</title>
    <link>https://answer.freetools.me/tags/post-norm/</link>
    <description>Recent content in Post-Norm on Answer</description>
    <generator>Hugo -- 0.152.2</generator>
    <language>zh-cn</language>
    <lastBuildDate>Wed, 11 Mar 2026 12:51:06 +0800</lastBuildDate>
    <atom:link href="https://answer.freetools.me/tags/post-norm/index.xml" rel="self" type="application/rss+xml" />
    <item>
      <title>残差连接：为什么 Transformer 能堆叠到百层而不梯度消失？</title>
      <link>https://answer.freetools.me/%E6%AE%8B%E5%B7%AE%E8%BF%9E%E6%8E%A5%E4%B8%BA%E4%BB%80%E4%B9%88-transformer-%E8%83%BD%E5%A0%86%E5%8F%A0%E5%88%B0%E7%99%BE%E5%B1%82%E8%80%8C%E4%B8%8D%E6%A2%AF%E5%BA%A6%E6%B6%88%E5%A4%B1/</link>
      <pubDate>Wed, 11 Mar 2026 12:51:06 +0800</pubDate>
      <guid>https://answer.freetools.me/%E6%AE%8B%E5%B7%AE%E8%BF%9E%E6%8E%A5%E4%B8%BA%E4%BB%80%E4%B9%88-transformer-%E8%83%BD%E5%A0%86%E5%8F%A0%E5%88%B0%E7%99%BE%E5%B1%82%E8%80%8C%E4%B8%8D%E6%A2%AF%E5%BA%A6%E6%B6%88%E5%A4%B1/</guid>
      <description>深入解析Transformer中残差连接的设计原理：从梯度消失问题到恒等映射的数学本质，从Pre-Norm与Post-Norm的权衡到DeepNet实现1000层训练，揭示这个让现代大模型成为可能的核心架构组件。</description>
    </item>
  </channel>
</rss>
