MATH
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
MATH
0.892ASR (Rate)
7
MATH 500
98.5Accuracy
7
MATH
92.84Accuracy (%)
7
MATH500
94Avg@4 Score
7
MATH (test)
64.5Algebra Score
7
MATH 500
70MATH 500 Score
7
MATH500
37.2Accuracy
7
MATH first 1,000 samples (test)
86.8Exact Match Accuracy
7
MATH 500
66.3Average@4
7
MATH Out-of-Domain
69.24LPM Score
7
MATH Out-of-Domain
73.9AUROC
7
MATH (test)
0.655ACC@5
7
MATH (eval)
115.29MAE
6
MATH LLaMA-2-7B (test)
19.8Accuracy
6
MATH 500 (Korean)
95.6Pass@1 Accuracy
6
MATH Hard
80.67Accuracy
6
MATH 500
81.6Avg@1 Score
6
MATH500
76.3Accuracy
6
MATH
43.9PCR
6
MATH
76.28Accuracy
6
MATH (in-distribution)
86.97HV Score
6
Math
0Frustration T20 (%)
6
MATH500 Base context
88.7Accuracy
6
MATH500
89.93Pass@1
6
MATH-Chat
91.5Accuracy
6