Unverified50% confidenceBenchmarkExact time
In Anthropic's tests, the failure rate of Naive RAG (the correct text chunk not in the top 20) was 5.7%, dropping to 3.7% with contextual embeddings, 2.9% with contextual BM25, and 1.9% with reranking added.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/18/2026
First Seen
Valid until: 1/16/2027
Sources
Related Entities
Related Claims
UnverifiedIn the 5-agent conversational cluster architecture (unstructured Markdown handoffs), the spec drift failure rate is 48.4%, the silent tool error propagation rate is 79.0%, the premature termination rate is 34.2%, and the end-to-end success rate is 18.5%.75% similarUnverifiedOn Track B, the flat RAG baseline can detect 76%–97% of true contradictions, but produces false positives on 16%–43% of surface-similar safe drafts (the specific ratio depends on the backend).70% similarUnverifiedAfter applying logit penalties, the number of reasoning tokens decreased by 11% to 19% across all precision formats.69% similarUnverifiedWhen the label error rate is below about 10%-20%, deep networks can still learn effective representations, and the degradation in generalization performance is limited.69% similarUnverifiedOnline A/B testing showed a 5% decrease in skip rate.69% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1063909API
curl https://kongchang.com/api/v1/knowledge/claims/1063909MCP
get_claim(id=1063909)