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

Commonsense Reasoning on HellaSwag (Accuracy, AUROC, ECE)

95.1Accuracy

Mashup Learning

23.23641.89360.5579.207Mar 16, 2024Jul 19, 2024Nov 22, 2024Mar 28, 2025Jul 31, 2025Dec 4, 2025Apr 9, 2026
Updated 8d ago

Evaluation Results

MethodLinks
2026.03
95.1--
2026.03
93.2--
2026.03
91.6--
2026.02
87.360.8570.102
2026.02
83.260.8270.016
2026.02
79.970.8040.181
2026.02
79.640.8090.182
2026.02
63.10.6980.357
2026.03
61.2--
2024.03
59.67--
2026.03
56.1--
2026.03
56.1--
2026.03
55--
2026.03
55--
2026.03
53.7--
2026.03
53.7--
2026.03
53.2--
2026.03
53.2--
2026.03
52.7--
2026.03
51.4--
2026.03
51.1--
2026.03
51--
2026.03
51--
2026.03
50.8--
2026.03
50.8--
2026.03
50.6--
2026.03
50.5--
2026.03
50--
2026.03
50--
2024.03
49.49--
2026.03
48.4--
2024.03
48.28--
2026.03
48--
2026.03
47.8--
2024.03
47.65--
2026.03
47.6--
2026.03
47.6--
2026.03
47.6--
2026.03
47.1--
2026.03
47.1--
2026.04
26.3--
2026.04
26.2--
2026.04
26.2--
2026.04
26.1--
2026.04
26.1--
2026.04
26--
2026.04
26--