Agents & knowledge
Retrieval-augmented generation
Generating answers using information retrieved at request time.
Also known as: RAG
Definition
RAG finds relevant material from an external collection and supplies it to a model as context for its response.
Why it matters
RAG can improve freshness and traceability, but retrieval quality, permissions and citation support must be evaluated independently.
Related concepts
- System under test
The complete configuration being evaluated—not only the model.
- Citation correctness
Whether a cited source genuinely supports the associated claim.
- Grounding
Connecting an output to supplied, authoritative evidence.