<?xml version="1.0" encoding="utf-8" standalone="yes"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
  <channel>
    <title>Pgvector on Aaron Wang的博客</title>
    <link>https://aarenwang.github.io/tags/pgvector/</link>
    <description>Recent content in Pgvector on Aaron Wang的博客</description>
    <generator>Hugo</generator>
    <language>zh-CN</language>
    <lastBuildDate>Fri, 22 May 2026 09:40:00 +0800</lastBuildDate>
    <atom:link href="https://aarenwang.github.io/tags/pgvector/index.xml" rel="self" type="application/rss+xml" />
    <item>
      <title>混合检索与 RAG 回答可信度：从召回 chunk 到带来源回答</title>
      <link>https://aarenwang.github.io/posts/agent/rag/rag-hybrid-retrieval-optimization/</link>
      <pubDate>Fri, 22 May 2026 09:40:00 +0800</pubDate>
      <guid>https://aarenwang.github.io/posts/agent/rag/rag-hybrid-retrieval-optimization/</guid>
      <description>围绕 RAG 检索和问答阶段，说明向量检索、关键词召回、RRF 排序、rerank、上下文构建、来源引用和 Chat log 如何提升回答可信度。</description>
    </item>
    <item>
      <title>从零搭一个中文电子书 RAG：为什么先做最小闭环</title>
      <link>https://aarenwang.github.io/posts/agent/rag/rag-system-architecture-design/</link>
      <pubDate>Tue, 19 May 2026 05:00:00 +0800</pubDate>
      <guid>https://aarenwang.github.io/posts/agent/rag/rag-system-architecture-design/</guid>
      <description>从中文电子书 RAG 的 MVP 入手，梳理上传、解析、切块、向量化、检索、问答和来源引用的最小闭环，以及第一版系统应该保留和暂缓的工程能力。</description>
    </item>
  </channel>
</rss>
