Omni-MATH
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
Omni-Math
0.91Pearson PCC
4
Omni-MATH Non-Reasoning episodes (human-annotated gold set)
89.34Accuracy
4
Omni-MATH
13ECE
3
Omni-MATH
35.4Accuracy (Best-of-64)
3
Omni-Math n=1876
0.82Pearson Correlation
2