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

Mathematical Reasoning on GSM8K (Acc, H100 TPS, TPF)

90.3Accuracy (%)

LightningRL

68.66874.28479.985.516Mar 4, 2026Mar 8, 2026Mar 13, 2026Mar 18, 2026Mar 22, 2026Mar 27, 2026Apr 1, 2026
Updated 16d ago

Evaluation Results

MethodLinks
2026.03
90.3336.035.58
2026.03
89.5--
2026.03
88.9105.552.85
2026.03
85--
2026.03
82--
2026.03
77.51502.21
2026.04
74.96.26-
2026.04
74.72.77-
2026.03
74.157.31
2026.04
73.22.88-
2026.03
73.1288.99.11
2026.04
73.19.11-
2026.03
72.6172.25.14
2026.04
72.61-
2026.04
72.65.14-
2026.03
69.5--