MATH
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
MATH
95.86Accuracy
114
MATH 500
96.6Pass@1 Rate
113
MATH
92.71Pass@1
112
MATH500 (full)
92.4Accuracy
111
MATH-500 (test)
91Accuracy
109
MATH 500
97.9MATH 500 Accuracy
106
MATH 500
6,749Average Tokens
104
MATH500
94.6Accuracy
104
MATH500
93Accuracy
96
MATH 500
100Accuracy (%)
94
MATH500
98.6Performance (%)
83
MATH500
96.2Accuracy
82
MATH-500
97.6Accuracy
79
MATH500
96.4Pass@1
77
MATH v1 (test)
83.8Accuracy
77
MATH (test)
80.9Accuracy
73
MATH 500
91.2Accuracy
73
Math
81.57Mean AUROC
72
MATH
10.86Speedup
68
MATH 500
98.5Pass@1
68
Math
0ASR
60
MATH500 (test)
61.2Exact Match Accuracy
60
MATH 500
60.6Exact Match
60
MATH 500
94.4Accuracy
60
MATH-500
97.4Pass@1 Accuracy
59