Confidence calibration
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
81.2AUROC
84
Feb 26, 2026
0.013Confidence Calibration Error
66
Mar 12, 2026
0.015ECE
53
Feb 26, 2026
0.001Calibration Error
44
Mar 12, 2026
0.008Calibration Error
44
Mar 12, 2026
1.1Calibration Error
44
Mar 12, 2026
3.1ECE
41
Jun 9, 2026
0.031ECE
36
Feb 26, 2026
0.031ECE
36
Feb 26, 2026
0.07ECE
28
Feb 26, 2026
0.039Brier Score
27
Feb 26, 2026
0.03Calibration Error
22
Mar 12, 2026
0.006Calibration Error
22
Mar 12, 2026
4.4ECE
21
May 13, 2026
8ECE
18
May 4, 2026
5.6ECE
18
May 4, 2026
3.4Expected Calibration Error (ECE)
18
May 4, 2026
0.114Brier Score
18
Feb 26, 2026
0.007ECE
16
Feb 26, 2026
1.5ECE
15
May 4, 2026
1.202SNR Gain
15
Feb 26, 2026
0.65ECE
12
Feb 26, 2026
0Brier Delta
10
May 20, 2026
3.26ECE
9
Feb 26, 2026
0.14ECE
9
Feb 26, 2026