List your company

Question Answering on AmbigQA (Accuracy, Correlation, and AUROC)

59.8AccuracyVariational Credal
57.40858.02958.6559.271Feb 11, 2026

Evaluation results

MethodLinks
Variational Credal2026.0259.80.080.350.420.71
Deep Ensembles2026.0259.10.80.270.210.6
CBDL2026.02590.610.310.270.64
CreINNs2026.0258.90.630.30.250.62
MC Dropout2026.0258.80.790.260.20.59
Semantic Entropy2026.0258.20.840.250.190.58
P(True)2026.0257.50.780.230.170.56