# Agentic RAG - [RAGで「AIエージェント」を使う手法まとめ](https://zenn.dev/knowledgesense/articles/64975fb9377f82) - [\[2501.09136\] Agentic Retrieval-Augmented Generation: A Survey on Agentic RAG](https://arxiv.org/abs/2501.09136) - 2025.01 - 設計パターン - Reflection(自己反省) - Planning(計画) - Tool Use(ツール利用) - Multi-Agent Collaboration(マルチエージェント協力) - 既存の RAG - Naïve RAG - Advanced RAG - Modular RAG - Graph RAG - Agentic RAG - アーキテクチャ - シングルエージェント型 - マルチエージェント型 - 階層型エージェント - Corrective RAG(セルフチェック付きRAG) - [RAGの新しい手法「CRAG」を3分で理解する](https://zenn.dev/knowledgesense/articles/bb5e15abb3c547) - [\[2401.15884\] Corrective Retrieval Augmented Generation](https://arxiv.org/abs/2401.15884) - Adaptive RAG(適応型RAG) - [RAGに質問分類させる「Adaptive-RAG」の解説](https://zenn.dev/knowledgesense/articles/8c23c35fa715c9) - [\[2403.14403\] Adaptive-RAG: Learning to Adapt Retrieval-Augmented Large Language Models through Question Complexity](https://arxiv.org/abs/2403.14403) - Graph-Based Agentic RAG - Agent-G - GeAR - [What is Agentic RAG | Weaviate](https://weaviate.io/blog/what-is-agentic-rag) - AI agent-based implementation of RAG - Architecture - Single-Agent RAG (Router) - Multi-agent RAG Systems - vs. (Vanilla) RAG - Access to external tools - Query pre-processing - Multi-step retrieval - Validation of retrieved information - Agent Frameworks - DSPy - LangChain - LlamaIndex - CrewAI - Swarm - Letta