<?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>Memory on Yuchen Du</title><link>https://dudusoar.github.io/tags/memory/</link><description>Recent content in Memory on Yuchen Du</description><generator>Hugo -- 0.164.0</generator><language>en-us</language><lastBuildDate>Fri, 17 Jul 2026 00:00:00 +0000</lastBuildDate><atom:link href="https://dudusoar.github.io/tags/memory/index.xml" rel="self" type="application/rss+xml"/><item><title>Separate Offline Memory Extraction from Online Action Execution</title><link>https://dudusoar.github.io/posts/llm-system-architecture/02-offline-memory-online-action/</link><pubDate>Fri, 17 Jul 2026 00:00:00 +0000</pubDate><guid>https://dudusoar.github.io/posts/llm-system-architecture/02-offline-memory-online-action/</guid><description>A clean LLM architecture separates the production of reusable experience from its read-only use during online decisions.</description></item><item><title>Establish Fact Ownership and Evidence Provenance in LLM Systems</title><link>https://dudusoar.github.io/posts/llm-system-architecture/08-fact-ownership-and-provenance/</link><pubDate>Fri, 17 Jul 2026 00:00:00 +0000</pubDate><guid>https://dudusoar.github.io/posts/llm-system-architecture/08-fact-ownership-and-provenance/</guid><description>Every fact, interpretation, proposal, and outcome in an LLM system needs a clear owner and a traceable evidence lineage.</description></item></channel></rss>