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

Mathematical Reasoning on Math500 (Pass@1, TPS, Latency)

93.5Pass@1 Acc

Target Model

70.176.17582.2588.325Feb 3, 2026Feb 10, 2026Feb 17, 2026Feb 24, 2026Mar 3, 2026Mar 10, 2026Mar 18, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
2026.02
93.540.7199.04
2026.02
92.839.2189.91
2026.02
92.1796.9543.53
2026.02
91117.3635.47
2026.03
90.7--
2026.03
90.4--
2026.03
90.2--
2026.03
89.8--
2026.02
89.552.6190.33
2026.03
89.1--
2026.02
8868.5859.32
2026.03
84.4--
2026.02
84149.1931.92
2026.02
82.3284.3415.94
2026.02
81.42380.5112.7
2026.02
81.42380.5112.7
2026.02
7776.0253.66
2026.02
71312.7727.99