<?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>Architecting Intelligence</title><link>https://blog.oklensuangoo.uk/</link><description>Recent content on Architecting Intelligence</description><generator>Hugo</generator><language>zh-cn</language><lastBuildDate>Sun, 15 Mar 2026 17:47:02 +0800</lastBuildDate><atom:link href="https://blog.oklensuangoo.uk/index.xml" rel="self" type="application/rss+xml"/><item><title>浅谈Agent的Memory发展</title><link>https://blog.oklensuangoo.uk/posts/%E6%B5%85%E8%B0%88agent%E7%9A%84memory%E5%8F%91%E5%B1%95/</link><pubDate>Sun, 15 Mar 2026 17:47:02 +0800</pubDate><guid>https://blog.oklensuangoo.uk/posts/%E6%B5%85%E8%B0%88agent%E7%9A%84memory%E5%8F%91%E5%B1%95/</guid><description>&lt;p&gt;随着LLM的地位完成了从语言模型到Agent 大脑的转变，LLM愈发需要进行长期持久性的上下文管理，这对原有的RAG和memory技术形成了新的挑战，本文将从该角度出发，阐述最近memory管理的最新研究进展&lt;/p&gt;</description></item><item><title>LLM 架构设计中的关键思考：我的第一篇博客</title><link>https://blog.oklensuangoo.uk/posts/first-post/</link><pubDate>Sun, 15 Mar 2026 15:15:00 +0800</pubDate><guid>https://blog.oklensuangoo.uk/posts/first-post/</guid><description>这篇文章记录了当前开源大模型架构选型过程中的心得，以及我个人博客搭建的初衷。</description></item><item><title>Memory 最新进展</title><link>https://blog.oklensuangoo.uk/posts/memory-%E6%9C%80%E6%96%B0%E8%BF%9B%E5%B1%95/</link><pubDate>Sun, 15 Mar 2026 15:15:00 +0800</pubDate><guid>https://blog.oklensuangoo.uk/posts/memory-%E6%9C%80%E6%96%B0%E8%BF%9B%E5%B1%95/</guid><description>这篇文章记录了当前大模型 Memory管理的最新进展</description></item><item><title/><link>https://blog.oklensuangoo.uk/posts//images/pasted/</link><pubDate>Mon, 01 Jan 0001 00:00:00 +0000</pubDate><guid>https://blog.oklensuangoo.uk/posts//images/pasted/</guid><description/></item><item><title/><link>https://blog.oklensuangoo.uk/posts/images/</link><pubDate>Mon, 01 Jan 0001 00:00:00 +0000</pubDate><guid>https://blog.oklensuangoo.uk/posts/images/</guid><description/></item><item><title/><link>https://blog.oklensuangoo.uk/posts/images/gemini_image.png/</link><pubDate>Mon, 01 Jan 0001 00:00:00 +0000</pubDate><guid>https://blog.oklensuangoo.uk/posts/images/gemini_image.png/</guid><description/></item><item><title/><link>https://blog.oklensuangoo.uk/posts/images/pasted/</link><pubDate>Mon, 01 Jan 0001 00:00:00 +0000</pubDate><guid>https://blog.oklensuangoo.uk/posts/images/pasted/</guid><description/></item><item><title/><link>https://blog.oklensuangoo.uk/posts/images/screenshot/</link><pubDate>Mon, 01 Jan 0001 00:00:00 +0000</pubDate><guid>https://blog.oklensuangoo.uk/posts/images/screenshot/</guid><description/></item><item><title>About Me</title><link>https://blog.oklensuangoo.uk/about/</link><pubDate>Mon, 01 Jan 0001 00:00:00 +0000</pubDate><guid>https://blog.oklensuangoo.uk/about/</guid><description>&lt;p&gt;你好！我是 Oklens，是一名资深大模型工程师。&lt;/p&gt;
&lt;p&gt;在这个博客中，我会记录和分享关于LLM架构与检索增强相关的学习新的。&lt;/p&gt;
&lt;p&gt;欢迎通过邮件与我交流探讨：&lt;a href="mailto:oklen@foxmail.com"&gt;oklen@foxmail.com&lt;/a&gt;&lt;/p&gt;</description></item></channel></rss>