Unverified50% confidenceDecision RuleExact time
Avoid questions like 'summarize this document'; instead use adversarial retrieval questions (mutually contradictory clauses, exceptions buried in appendices, numbers that appear only in tables) to evaluate RAG retrieval quality.
1
Sources
50%
Confidence
Long-term
Relevance
10/14/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedChunks that are too large dilute retrieval precision, while chunks that are too small result in incomplete semantics; the chunking strategy directly determines the retrieval quality of a RAG system.68% similarUnverifiedTraditional RAG performs very poorly on tasks that require globally scattered evidence; retrieving a few relevant text chunks and analyzing an entire document collection are two different tasks.68% similarUnverifiedCommon RAG mistakes include improper chunking strategies, mismatch between the embedding model and queries, and lack of reranking on retrieved results.66% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1031947API
curl https://kongchang.com/api/v1/knowledge/claims/1031947MCP
get_claim(id=1031947)