Unverified50% confidenceFactExact time
The benchmark produces two numbers for each model that are never merged: a task score and a false-confidence rate (the proportion of cases that should have been escalated but were answered anyway).
1
Sources
50%
Confidence
Long-term
Relevance
10/12/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedA single performance percentage is misleading; a model that claims to retain 98% of performance may happen to fall within the lost 2% for a specific task type, and real tasks should be tested before trusting benchmarks.78% similarUnverified对同一测试用例执行多次调用统计成功比例得到通过率，比二元的通过/失败判断更能反映模型真实表现76% similarUnverifiedRanked 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.75% similarUnverifiedTwo models with similar intelligence scores may differ in inference cost by several times or even an order of magnitude.73% similarUnverifiedBenchmark scores measure the average of a model's success consistency and failure frequency, not its peak capability.71% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1002960API
curl https://kongchang.com/api/v1/knowledge/claims/1002960MCP
get_claim(id=1002960)