Unverified50% confidenceFactExact time
When evaluating a judge, one should look at the confusion matrix rather than just accuracy; if 90% of the samples are good answers to begin with, a judge that passes everything can still show 90% accuracy.
1
Sources
50%
Confidence
Long-term
Relevance
10/19/2026
First Seen
Sources
Related Entities
Related Claims
Unverified校准的不确定性区间意味着：若模型声称 90% 置信区间为某范围，则在大量测试样本中约 90% 的真实值应落入该范围。76% similarVerifiedCalibrated confidence means that when the system gives a score of 0.9, the probability of being correct should be close to 90%.75% similarUnverified当所有模型在某评测上都能拿90分时该评测就应退场，即基准饱和现象75% similarUnverifiedOn reasoning-based question-answering tasks, the strongest closed-source models achieved accuracy exceeding 90%.74% similarUnverifiedThe 98% score from a single classification is the confidence level for that message's corresponding category, and does not represent an overall model accuracy of 98%.74% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1066493API
curl https://kongchang.com/api/v1/knowledge/claims/1066493MCP
get_claim(id=1066493)