<?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>Context-Compaction on CYR&#39;ML</title>
    <link>https://chengyongru.github.io/tags/context-compaction/</link>
    <description>Recent content in Context-Compaction on CYR&#39;ML</description>
    <generator>Hugo</generator>
    <language>en-us</language>
    <lastBuildDate>Mon, 07 Sep 2026 18:13:07 +0800</lastBuildDate>
    <atom:link href="https://chengyongru.github.io/tags/context-compaction/index.xml" rel="self" type="application/rss+xml" />
    <item>
      <title>nanobot 的上下文压缩和记忆机制</title>
      <link>https://chengyongru.github.io/blog/notebook/nanobot%E4%B8%8A%E4%B8%8B%E6%96%87%E9%87%8D%E6%9E%84-%E4%BA%94%E6%9D%A1%E7%94%9F%E5%91%BD%E5%91%A8%E6%9C%9F/</link>
      <pubDate>Mon, 07 Sep 2026 17:51:25 +0800</pubDate>
      <guid>https://chengyongru.github.io/blog/notebook/nanobot%E4%B8%8A%E4%B8%8B%E6%96%87%E9%87%8D%E6%9E%84-%E4%BA%94%E6%9D%A1%E7%94%9F%E5%91%BD%E5%91%A8%E6%9C%9F/</guid>
      <description>&lt;p&gt;最近对 nanobot 的上下文压缩机制做了不少改动, 在这里稍作记录, 也顺便聊聊 nanobot 现在的长期记忆机制, 希望能对各位有所帮助, 如有错漏, 也欢迎批评指正.&lt;/p&gt;
&lt;h2 id=&#34;上下文怎么压缩&#34;&gt;上下文怎么压缩&lt;/h2&gt;
&lt;p&gt;不严谨地讲, 上下文可以看成一个消息列表, 大致长这样:&lt;/p&gt;</description>
    </item>
  </channel>
</rss>
