easy5 pts · first answer countsragvector-databases

Why store metadata with each chunk

Why do RAG pipelines store fields like source URL, title, department and last-updated date alongside each chunk's vector?

A.To make the vectors smaller so the index uses less memoryB.Because the embedding model needs them to compute the vectorC.To stop the LLM from reading the chunk textD.To filter searches, show citations and handle updates to the source

Discussion 0

Join free to post in the discussion.

Why store metadata with each chunk

Discussion 0

Related questions