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

Multitask Language Understanding on MMLU (Acc, AUROC, ECE)

71.4Accuracy

C3RL (Ours)

36.0445.2254.463.58Jul 2, 2026
Updated 23d ago

Evaluation Results

MethodLinks
2026.07
71.40.6530.092
2026.07
71.20.6480.111
2026.07
71.10.6020.239
2026.07
70.70.6070.175
2026.07
69.90.7590.141
2026.07
68.30.5870.188
2026.07
640.7310.129
2026.07
640.680.089
2026.07
63.50.6880.087
2026.07
63.40.5370.208
2026.07
62.30.670.153
2026.07
62.10.7060.098
2026.07
59.90.7330.158
2026.07
59.60.5860.271
2026.07
52.80.6280.162
2026.07
37.40.7150.11