Retrieval-Augmented Generation (RAG)(Retrieval-Augmented Generation)

Retrieval-Augmented Generation (RAG) integrates external knowledge retrieval into the LLM generation process.

Overview

For a given user query, RAG first retrieves relevant documents, then includes them in the prompt sent to the LLM. This mitigates training-time knowledge cutoffs and reduces hallucinations.

Key components

• Retriever: Fetches relevant documents from a vector DB / search engine
• Generator: An LLM that produces the final answer given retrieved context and the query
• Embeddings: Maps documents and queries into a shared vector space