Definition
A method in which a language model retrieves relevant, current documents at the time of a query and conditions its answer on them, instead of answering from its training alone.
How we use the term
Generative search is built on it; the quality of the answer is bounded by what the retrieval finds.
Also called
RAG
Where it sits
Broader
On this site
Elsewhere
WikidataHW-CON-064
All concepts