Unverified50% confidenceFactExact time
The model showed the highest confidence (95%) on underdetermined problems (problems with no unique correct answer), exhibiting a reverse calibration phenomenon.
1
Sources
50%
Confidence
Long-term
Relevance
10/12/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedRanked by question difficulty, persistence rate was 91.4% on the 60 questions that all models answered correctly, but only 73.5% on the 52 questions that at least one model got wrong, indicating that sycophancy tracks uncertainty.78% similarUnverifiedMany models still exhibit high confidence in incorrect answers, presenting a confidence calibration problem.78% similarUnverifiedAnthropic的研究显示，即使面对毫无根据的质疑，在某些测试中模型被用户简单反驳后放弃正确答案的比例超过50%77% similarVerifiedIn scenarios with extreme class imbalance, even an accuracy as high as 95% may indicate that the model is completely unable to identify the minority class.76% similarUnverified仅用小模型策略的每正确答案成本最低($0.0004)但准确率仅33%，说明没有准确率底线的每正确答案成本指标具有误导性76% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1005672API
curl https://kongchang.com/api/v1/knowledge/claims/1005672MCP
get_claim(id=1005672)