MATH
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
MATH 500
58.2Multi@5 Accuracy
16
MATH 500
83.3Accuracy (avg@32)
16
Math_500 CodaSet OOD (test)
90.33Accuracy (%)
16
MATH 500
99.97LLM Trust Score
16
MATH original and paraphrased (train)
35.29Reference Score/Rate
16
MATH 500
4.47Success Rate (SR)
16
MATH
46.1Level
16
MATH
80.2Mean Harmonic Score (MH)
16
MATH 500
95.8M500 Score
16
MATH 500
83.6Accuracy
16
MATH 500
94.2Pass@1 Accuracy
16
math
0.037MAE
16
MATH (test)
70.6Accuracy (r*)
16
MATH
97.09Coverage
16
MATH
90.8Pass@1
16
MATH Camel
0.735AUROC
16
MATH
0.712AUROC
16
MATH AutoGen (test)
0.754AUROC
16
MATH 500
89.16Accuracy
16
MATH 500
85.2Average Success Rate @32
16
MATH 500
108.5Retention Rate
16
MATH
76.8Accuracy@16
16
Math (val)
100Pass@16
16
MATH500
66.2Pass@1
16
Math-500
13.86Throughput
16