<?xml version="1.0" encoding="utf-8" standalone="yes"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
  <channel>
    <title>Hallucination on CATNAP</title>
    <link>/tags/hallucination/</link>
    <description>Recent content in Hallucination on CATNAP</description>
    <generator>Hugo -- gohugo.io</generator>
    <language>zh-TW</language>
    <copyright>© 2026 Idle Engineer</copyright>
    <lastBuildDate>Fri, 08 May 2026 00:00:00 +0000</lastBuildDate><atom:link href="/tags/hallucination/index.xml" rel="self" type="application/rss+xml" />
    
    <item>
      <title>LLM 幻覺是什麼｜為什麼 AI 會說謊與如何降低</title>
      <link>/ai/llm-hallucination/</link>
      <pubDate>Fri, 08 May 2026 00:00:00 +0000</pubDate>
      
      <guid>/ai/llm-hallucination/</guid>
      <description>幻覺（Hallucination）是 LLM 生成看起來合理但實際上錯誤的內容。根本原因是模型在預測機率，不在查事實。RAG、自我核實、降低 Temperature 是主要緩解手段。</description>
      
    </item>
    
  </channel>
</rss>
