RAG

Also known as: Retrieval-Augmented Generation

Retrieval-augmented generation (RAG) retrieves relevant documents at query time and conditions the LLM on that context, instead of relying only on parametric memory.

Typical pipeline: chunk corpus → embed → index → retrieve top-$k$ → (optional rerank) → prompt the model with passages. RAG improves factual grounding and updatability; CPT/SFT still help the model use retrieved text better.

← Back to AI Glossary