Loading video player...
A short podcast-style discussion on how to evaluate hybrid retrieval in RAG systems, including why better retrieval scores do not always mean better answers, how sparse and dense signals behave differently, and why query-type breakdowns matter.