Grounding
Tying a model’s answer to specific supplied sources, such as retrieved documents or search results, instead of its training memory.
A grounded answer is built from text placed in the context: documents found by RAG, database rows, web search results. The prompt tells the model to answer only from that material and to say when the answer is not there. Grounding reduces hallucination and makes answers checkable, but only as far as the sources are right and the retrieval found them. Faithfulness is the metric that measures whether it held.
Related terms
- RAG (retrieval-augmented generation): Finding relevant documents first and putting them into the prompt, so the model answers from them.
- Hallucination: A fluent, confident answer that is false or not supported by the sources.
- Faithfulness (groundedness): A RAG metric: the share of claims in an answer that are supported by the retrieved sources, not by the model’s memory or invention.
- Citations (source attribution): Links from each claim in an AI answer to the passage it came from, so a reader can verify it.
