Retrieval-augmented generation
Section titled “Retrieval-augmented generation”Retrieval-augmented generation, or RAG, retrieves external information relevant to a request and supplies that information to a generative model.
Retrieval can improve freshness, provenance, and task-specific grounding without changing the model’s learned parameters.
RAG does not guarantee correctness. Retrieval quality, context construction, and answer evaluation remain separate engineering concerns.
Related knowledge
Retrieval-augmented generationRetrieve external knowledge for a task and provide the selected evidence to the model at generation time.