Unverified50% confidenceBenchmarkExact time
On 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).
1
Sources
50%
Confidence
Long-term
Relevance
9/25/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedAn accuracy rate of 86% means there is still a 14% margin of error, making Articos more suitable as an early validation and directional judgment tool rather than a replacement for real user testing.69% similarUnverifiedOn a test subset of 152 completely untouched questions, anchored hybrid retrieval achieved 42.11%, while the planner-guided method reached only 36.84%.67% similarUnverified报告同时给出中位数KLD和99.9%分位KLD两个数值，前者反映典型情况，后者捕捉极端长尾错误67% similarUnverifiedAggregate metrics (such as average scores) mask failures in long-tail scenarios; when 95% of routine cases perform excellently, the collapse of the remaining 5% of edge cases leaves no obvious mark on the overall score.67% similarUnverifiedThe DeepSuite benchmark has a false positive rate of 44.9% and a false negative rate of 1.2%.67% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/949060API
curl https://kongchang.com/api/v1/knowledge/claims/949060MCP
get_claim(id=949060)