QuestionQ47

Verified

An agent uses policy documents retrieved from a knowledge base to compose an answer. You need to verify that the generated answer strictly reflects the retrieved content and introduces no unsupported statements.

Which evaluator should you configure?

Explanation

An LLM-as-a-judge faithfulness evaluator checks whether the claims in a generated answer are supported by the retrieved policy documents. It targets unsupported statements rather than just matching wording or comparing overall meaning. (learn.microsoft.com)

Learn more

Community Discussion

No comments yet. Be the first to start the discussion!