Unverified50% confidenceFactExact time
Error dependence is stronger among high-accuracy frontier model reviewers, even when those reviewers come from different vendors.
1
Sources
50%
Confidence
Long-term
Relevance
9/23/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedDenser relevance judgments reduce the probability that correct results retrieved by a model are mistakenly judged as irrelevant, thereby improving evaluation reliability.78% similarUnverifiedErrors consistent with the recipient's inclination degrade model performance more severely than random errors with an equivalent error rate.77% similarUnverified提高排序精度往往意味着模型更依赖接近事件发生时的强信号特征，可能牺牲早期预警提前量，这种权衡在医疗诊断和金融欺诈检测领域同样存在74% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/942778API
curl https://kongchang.com/api/v1/knowledge/claims/942778MCP
get_claim(id=942778)