MATH
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
MATH eval (test)
59.1Success Rate
20
MATH Minerva
28.3Pass@8 Rate
20
MATH
0FPR (%)
20
MATH (test)
84.27MATH Accuracy
20
MATH level 5
84Success Rate
20
Math500
20,417Throughput (tok/s)
20
MATH
38.24Accuracy
20
Math Base-9
92.7Accuracy
20
MATH
33.6Accuracy
20
MATH
0.362Overall Accuracy
20
Math-500
94.34Pass@1 Accuracy
19
MATH
88.3Pass@1
19
MATH 500
91.6Arithmetic Accuracy
19
MATH-500
81.6Accuracy
19
MATH 500
94.6Accuracy
19
MATH
46.2Pass@1
19
MATH500
76Pass@3 Accuracy
19
MATH500 0-shot
47.4Pass@1
19
MATH 500
45.19Pass@1
18
MATH500
84.6Mean@8 Score
18
MATH500
18.48BCE
18
MATH 500
88.5Accuracy
18
MATH-500 latest (held-out)
89.4Pass@1 Score
18
MATH-500
81.6Accuracy (MATH-500)
18
MATH
70PRR
18