<rss xmlns:atom="http://www.w3.org/2005/Atom" version="2.0">
    <channel>
        <title>MiniMax - 标签 - vllbc02&#39;s blogs</title>
        <link>https://blog.vllbc.top/tags/minimax/</link>
        <description>MiniMax - 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/minimax/" rel="self" type="application/rss+xml" /><item>
    <title>MiniMax M2</title>
    <link>https://blog.vllbc.top/minimax-m2/</link>
    <pubDate>Sat, 06 Jun 2026 00:00:00 &#43;0000</pubDate>
    <author>vllbc</author>
    <guid>https://blog.vllbc.top/minimax-m2/</guid>
    <description><![CDATA[<h1 id="minimax-m2">MiniMax M2</h1>
<blockquote>
<p>原文第 13 节。MiniMax（MiniMax）在 M1 之后发布的旗舰。benchmark
表现目前被认为是开源 SOTA。</p>
</blockquote>
<h2 id="关键背景">关键背景</h2>
<ul>
<li>M1（2025-06）用了 lightning attention（线性 attention 变体），456B
总参 / 46B 激活</li>
<li>M2（2025-12 之后）<strong>回退到 full attention</strong>——团队公开说
linear attention 在生产 LLM 里 tricky，普通 prompt 看起来 OK，但
reasoning / multi-turn / agent 掉点严重</li>
</ul>
<p></p>]]></description>
</item>
</channel>
</rss>
