<rss xmlns:atom="http://www.w3.org/2005/Atom" version="2.0">
    <channel>
        <title>Attention Sinks - 标签 - vllbc02&#39;s blogs</title>
        <link>https://blog.vllbc.top/tags/attention-sinks/</link>
        <description>Attention Sinks - 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/attention-sinks/" rel="self" type="application/rss+xml" /><item>
    <title>Attention Sinks</title>
    <link>https://blog.vllbc.top/attention-sinks/</link>
    <pubDate>Sat, 06 Jun 2026 00:00:00 &#43;0000</pubDate>
    <author>vllbc</author>
    <guid>https://blog.vllbc.top/attention-sinks/</guid>
    <description><![CDATA[<h1 id="attention-sinks">Attention Sinks</h1>
<blockquote>
<p>来自 Sebastian Raschka 的 <em>The Big LLM Architecture
Comparison</em>。</p>
</blockquote>
<h2 id="一句话">一句话</h2>
<p>在序列开头放一个”永远被 attend 到”的特殊 token，让长上下文 attention
稳定。这个特殊 token 会自然学到”整序列通用信息”。</p>
<h2 id="起源">起源</h2>
<p><em>Efficient Streaming Language Models with Attention Sinks</em>
论文。</p>]]></description>
</item>
<item>
    <title>GPT-OSS（OpenAI 的开源模型）</title>
    <link>https://blog.vllbc.top/gpt-oss/</link>
    <pubDate>Sat, 06 Jun 2026 00:00:00 &#43;0000</pubDate>
    <author>vllbc</author>
    <guid>https://blog.vllbc.top/gpt-oss/</guid>
    <description><![CDATA[<h1 id="gpt-oss">GPT-OSS</h1>
<blockquote>
<p>原文第 9 节。OpenAI 自
GPT-2（2019）以来首次发布开源模型，2025-08。两个 size：gpt-oss-20b 和
gpt-oss-120b。</p>
</blockquote>
<h2 id="总体架构">总体架构</h2>
<p>整体骨架”很眼熟”——所有组件都在其他模型中见过。3.6B 激活参数，MoE +
滑动窗口 attention + GQA。</p>]]></description>
</item>
</channel>
</rss>
