Share your thoughts, 1 month free Claude Pro on usSee more

Open-ended Multimodal Understanding Benchmark Average

0.799AUC

COMET

Updated 24d ago

Evaluation Results

Method	Links
COMET 2026.06		0.799	0.728	0.701
COMET 2026.06		0.785	0.784	0.707
COMET 2026.06		0.784	0.631	0.717
COMET 2026.06		0.734	0.647	0.771
P(Correct) 2026.06		0.714	0.652	0.766
P(Correct) 2026.06		0.712	0.536	0.837
P(Correct) 2026.06		0.685	0.648	0.863
P(Correct) 2026.06		0.667	0.532	0.897
Consist. (Ent) 2026.06		0.661	0.464	0.899
Consist. (Max) 2026.06		0.66	0.464	0.899
Consist. (Ent) 2026.06		0.657	0.548	0.904
Consist. (Max) 2026.06		0.655	0.547	0.904
VL Uncertainty 2026.06		0.654	0.605	0.829
VL Uncertainty 2026.06		0.654	0.665	0.829
VL Uncertainty 2026.06		0.65	0.646	0.852
Verbalized Conf 2026.06		0.642	0.598	0.787
Verbalized Conf 2026.06		0.639	0.546	0.84
VL Uncertainty 2026.06		0.637	0.535	0.869
Consist. (Ent) 2026.06		0.624	0.593	0.918
Consist. (Max) 2026.06		0.623	0.593	0.914
Verbalized Conf 2026.06		0.591	0.592	0.842
NLL 2026.06		0.576	0.433	0.949
NLL 2026.06		0.569	0.545	0.963
Verbalized Conf 2026.06		0.562	0.378	0.896
Consist. (Ent) 2026.06		0.495	0.411	0.959
Consist. (Max) 2026.06		0.495	0.411	0.959
NLL 2026.06		0.491	0.357	0.974
NLL 2026.06		0.475	0.417	0.97