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

Zero-shot Language Understanding on ARC-E, ARC-C, HellaSwag, LAMBADA, and PIQA (lm-eval 0.4.11 test)

81.5Average Accuracy

BF16

60.766.171.576.9Mar 30, 2026
Updated 19d ago

Evaluation Results

MethodLinks
2026.03
81.5
2026.03
80.9
2026.03
80.9
2026.03
80.8
2026.03
80.8
2026.03
79.5
2026.03
76.4
2026.03
76.4
2026.03
76
2026.03
76
2026.03
75.9
2026.03
75.7
2026.03
75.5
2026.03
75.5
2026.03
75.3
2026.03
75.2
2026.03
74.9
2026.03
74.6
2026.03
74.6
2026.03
74.6
2026.03
74.2
2026.03
74
2026.03
73.9
2026.03
72.8
2026.03
70.5
2026.03
70
2026.03
69.4
2026.03
69.3
2026.03
69.2
2026.03
68.7
2026.03
68.4
2026.03
67.6
2026.03
67.5
2026.03
67
2026.03
66.3
2026.03
65.1
2026.03
64.3
2026.03
64.1
2026.03
63.9
2026.03
63.8
2026.03
63.1
2026.03
61.5