Before releasing a knowledge-heavy agent, the QA lead wants evidence answers stick to sources rather than inventing claims. Which evaluation focus is most relevant?
Select an answer to reveal the explanation.
Short Explanation and Infographic
Groundedness asks: did the answer stick to the retrieved sources? Brand colors and font sizes don’t answer that — test questions and citation checks do.
Full explanation below image
Full Explanation
Agent evaluation for RAG systems emphasizes groundedness, relevance, and related quality measures. Test sets should include questions with known source answers and adversarial gaps.
Option A is correct.
Options B–D are irrelevant vanity metrics.
Exam tip: Groundedness ≠ relevance (support vs topical fit); know both terms.