MATH
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
Math500
53.2Accuracy
10
MATH Hard
84.4Ac (alpha=0.10)
10
MATH 500
84.2Expected Utility
10
MATH500
32.8Accuracy
10
Math 500
98.44Pass@4 Success Rate
10
Math 500
72.03Accuracy
10
Math GSM8K MATH
67Accuracy
10
MATH (test)
4Runtime (s)
10
MATH
76.67Accuracy (MATH)
10
MATH500
89.8Average Score (Avg@32)
10
MATH500
91.4Pass@16
10
MATH 500
84.68Avg Success Rate (@8)
10
MATH500
93.6pass@4
10
MATH-500
96.4Score
10
Hendrycks-MATH
1.2Average Acceptance Length (τ)
10
MATH
43.2Accuracy
10
MATH-500
52.94Correction Uplift (%)
10
MATH
48.87Correction Uplift
10
MATH
99.9AUC (Unspecified Config)
10
Math500
81.6Pass@8
10
MATH (test)
39.8Accuracy
10
MATH
39.6Accuracy
10
MATH500
97.5Accuracy (Avg@4)
10
MATH 500
79.95Score
10
MATH
36.9Accuracy
10