Categories Companies Research

Loading the SOTA2 catalog…

Explore

Categories Companies Research

Contribute

Join SOTA2 Contact

SOTA2

© 2026 SOTA2

Build in Bellevue, USA

WizWand is now SOTA2 Research

Papers Benchmarks Datasets Tasks

Probability Calibration research benchmarks · SOTA2 Research

Probability Calibration

Benchmarks

Dataset NameSOTA MethodSort

Dataset NameSOTA methodMetricTrendResultsLast Updated

ChaosNLI combined SNLI+MNLI K=3

2.15ECE

26

Mar 25, 2026

CIFAR-10H K=10 (test)

0.7ECE

26

Mar 25, 2026

HalLocalizer (VisualBERT)

4.28Object ECE

12

Feb 26, 2026

HalLoc-Instruct

HalLocalizer (InternVL + VisualBERT)

0.11Object ECE

12

Feb 26, 2026

HalLoc-Caption (test)

HalLocalizer (InternVL + VisualBERT)

0.04Object ECE

12

Feb 26, 2026

ImageNet 16H Noise Level 125

719ECE

7

Feb 26, 2026

ImageNet-16H Noise Level 110

3.92ECE

7

Feb 26, 2026

ImageNet 16H Noise Level 95

3.23ECE

7

Feb 26, 2026

Obesity dataset (Xh)

0.014MCLLO p-value

4

Feb 26, 2026

Obesity dataset (Xt)

0.16ECE (sqrt(n) bins)

1

Feb 26, 2026

Popular datasets

ImageNet CIFAR Obesity ChaosNLI HalLoc

Follow for updates

Join SOTA2 Contact