整理文件路径:Technical→AI/
This commit is contained in:
37
wiki/sources/llms-rag-ai-agent-qu-bie.md
Normal file
37
wiki/sources/llms-rag-ai-agent-qu-bie.md
Normal file
@@ -0,0 +1,37 @@
|
||||
---
|
||||
title: "LLMs、RAG、AI Agent三个到底什么区别"
|
||||
type: source
|
||||
tags: [llm, rag, ai-agent]
|
||||
date: 2025-11-19
|
||||
---
|
||||
|
||||
## Source File
|
||||
- [[raw/AI/LLMs、RAG、AI Agent 三个到底什么区别?.md]]
|
||||
|
||||
## Summary
|
||||
- 核心主题:LLM、RAG、AI Agent三者区别
|
||||
- 问题域:AI应用开发初学者容易混淆这三个概念
|
||||
- 方法/机制:分层定位,LLM思考、RAG认知、Agent执行
|
||||
- 结论/价值:三者不是竞争技术,而是在不同层面满足不同场景的能力展示
|
||||
|
||||
## Key Claims
|
||||
- LLM是"天才大脑",擅长思考但对当前情况一无所知
|
||||
- RAG是"随身图书馆助理",提供实时信息消除幻觉
|
||||
- AI Agent是智能体,围绕LLM构建循环控制系统,能感知、规划、执行、反思
|
||||
- 未来架构设计在于将三者结合:LLM推理、RAG准确性、Agent自主性
|
||||
|
||||
## Key Concepts
|
||||
- [[LLM]]:大语言模型,AI应用的"天才大脑"
|
||||
- [[RAG]]:检索增强生成,AI的"随身图书馆助理"
|
||||
- [[AI Agent]]:智能体,具备感知-规划-执行-反思的循环控制
|
||||
- [[幻觉]]:LLM生成错误信息的问题
|
||||
|
||||
## Key Entities
|
||||
|
||||
## Connections
|
||||
- [[LLM]] ← provides ← [[思考能力]]
|
||||
- [[RAG]] ← extends ← [[LLM]]
|
||||
- [[AI Agent]] ← uses ← [[LLM]]
|
||||
- [[AI Agent]] ← uses ← [[RAG]]
|
||||
|
||||
## Contradictions
|
||||
Reference in New Issue
Block a user