Unverified50% confidenceFactExact time
When large models answer directly, their pass rate on implicit scoring criteria is only 67%, while their pass rate on explicit criteria reaches 81%.
1
Sources
50%
Confidence
Long-term
Relevance
10/27/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedAt the mention level, the model achieved an overall precision of 74.1% and a recall of 70.5%.80% similarUnverifiedOn a 50-sample evaluation set, an accuracy change from 74% to 76% corresponds to just one more correct sample, a difference that is statistically almost meaningless.76% similarUnverifiedOn the full validation set, EMBER achieves 73.5% accuracy and ORBIT 71.2%, while the majority-class baseline is only 51.5%.75% similarUnverifiedWhen the model claims at least 90% confidence, its actual accuracy is only 56%, making it an extremely poorly calibrated model; therefore ReachNT hides the confidence percentage from tenants, and 42% of dangerous reports are routed to human handling before any decision is made.74% similarUnverifiedAcross 79 pairs, the cumulative evasion rate reached 40.5% under the detailed feedback condition and 20.3% under the generic rejection condition.73% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1102085API
curl https://kongchang.com/api/v1/knowledge/claims/1102085MCP
get_claim(id=1102085)