AIME
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
AIME 24
0.941Mean Kendall's tau_b
1
AIME 25
0.798Kendall's tau_b (vs Gold Standard)
1
AIME '24
0.779Kendall's tau_b (vs. Gold)
1
AIME VeRA-H / VeRA-H Pro 2024-II
0.849Avg@5 Accuracy (Seeds)
1
AIME VeRA-E 2025
79.21Avg@5 Acc (Seeds)
1
AIME VeRA-E 2024
84.46Avg Accuracy @5 (Seeds)
1
AIME 24
—Acc@16
0
AIME25
—Output Tokens
0
AIME 2024
—Primary metric
0