<rss xmlns:atom="http://www.w3.org/2005/Atom" version="2.0">
    <channel>
        <title>Pre-Norm - 标签 - vllbc02&#39;s blogs</title>
        <link>https://blog.vllbc.top/tags/pre-norm/</link>
        <description>Pre-Norm - 标签 - vllbc02&#39;s blogs</description>
        <generator>Hugo -- gohugo.io</generator><language>zh-CN</language><copyright>This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.</copyright><lastBuildDate>Sat, 06 Jun 2026 00:00:00 &#43;0000</lastBuildDate><atom:link href="https://blog.vllbc.top/tags/pre-norm/" rel="self" type="application/rss+xml" /><item>
    <title>Post-Norm vs Pre-Norm</title>
    <link>https://blog.vllbc.top/post-norm-vs-pre-norm/</link>
    <pubDate>Sat, 06 Jun 2026 00:00:00 &#43;0000</pubDate>
    <author>vllbc</author>
    <guid>https://blog.vllbc.top/post-norm-vs-pre-norm/</guid>
    <description><![CDATA[<h1 id="post-norm-vs-pre-norm">Post-Norm vs Pre-Norm</h1>
<blockquote>
<p>来自 Sebastian Raschka 的 <em>The Big LLM Architecture
Comparison</em>。汇总 2025 年 LLM 中 normalization 放置的各种设计。</p>
</blockquote>
<h2 id="一句话">一句话</h2>
<p>Normalization 放在 attention / FFN
之前还是之后，对训练稳定性影响很大。2025 年的 LLM 至少分 4 派。</p>]]></description>
</item>
</channel>
</rss>
