MATH
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
MATH500
43.2Accuracy
12
MATH 500
89Accuracy
12
Math500
74Acc@16
12
MATH
83.3Perturbed SR
12
Math (test)
98.8AuROC
12
MATH
85.92Pass@1
12
MATH-500
80.2Pass@1
12
MATH
4.9Speedup
12
MATH500
0.93pass@1
12
MATH 500
76.2Avg@3
12
MATH 500
96Pass@1
12
MATH
4.46Accuracy
12
MATH
4.56Performance (%)
12
MATH level 3
56Accuracy
12
MATH-Perturb MP-hard
56.3Exact Match Accuracy
12
MATH 500
6.09Speedup
12
MATH (test)
53.6Accuracy
12
MATH-500
90Accuracy
12
Math latest (test)
20AIME2025
12
Math
75.82Accuracy
12
MATH Subset
44.2Accuracy
12
MATH
2.3Algebra Accuracy
12
Math500
88.8Accuracy (Math500)
11
MATH
76.12Accuracy
11
MATH Algebra (test)
98.15Accuracy
11