List your company

Math Reasoning on AIME 2024 (Accuracy, Avg, Delta)

91.25AccuracyFull
-3.6520.987545.62570.2625May 23, 2025

Evaluation results

MethodLinks
FullBase Model=Qwen3-30B-A3B, Sparsity=50%2025.0591.2580.69—
PreMoEBase Model=Qwen3-30B-A3B, Sparsity=50%2025.0588.3379.92-0.77
FullBase Model=openPangu-Ultra, Sparsity=31.25%2025.0580.8379.25—
PreMoEBase Model=openPangu-Ultra, Sparsity=31.25%2025.0580.4178.45-0.8
PreMoEBase Model=DeepSeek-R1, Sparsity=50%2025.0579.5876.521.01
EASY-EPBase Model=DeepSeek-R1, Sparsity=50%2025.0579.1774.69-0.82
Act-LogitsBase Model=openPangu-Ultra, Sparsity=31.25%2025.0577.576.91-2.34
EASY-EPBase Model=openPangu-Ultra, Sparsity=31.25%2025.0577.276.45-2.8
FullBase Model=DeepSeek-R1, Sparsity=50%2025.0577.0875.51—
FrequencyBase Model=openPangu-Ultra, Sparsity=31.25%2025.0576.2569.67-9.58
SEER (L)Base Model=openPangu-Ultra, Sparsity=31.25%2025.0575.4171.29-7.96
SEER (G)Base Model=openPangu-Ultra, Sparsity=31.25%2025.0571.6770.93-8.32
Act-LogitsBase Model=DeepSeek-R1, Sparsity=50%2025.057062.87-12.64
EASY-EPBase Model=Qwen3-30B-A3B, Sparsity=50%2025.0562.4654.26-26.43
FrequencyBase Model=DeepSeek-R1, Sparsity=50%2025.0560.8349.93-25.58
FrequencyBase Model=Qwen3-30B-A3B, Sparsity=50%2025.0560.4254.26-32.76
RandomBase Model=openPangu-Ultra, Sparsity=31.25%2025.0541.2550.13-29.12
RandomBase Model=DeepSeek-R1, Sparsity=50%2025.056.6726.69-48.82
RandomBase Model=Qwen3-30B-A3B, Sparsity=50%2025.05513.07-67.62
SEER (G)Base Model=DeepSeek-R1, Sparsity=50%2025.053.3324.31-51.2
SEER (L)Base Model=DeepSeek-R1, Sparsity=50%2025.052.9123.56-51.95
All-LogitsBase Model=DeepSeek-R1, Sparsity=50%2025.0505.46-70.05
All-LogitsBase Model=openPangu-Ultra, Sparsity=31.25%2025.0505.39-73.86
All-LogitsBase Model=Qwen3-30B-A3B, Sparsity=50%2025.0500.35-80.34
SEER (L)Base Model=Qwen3-30B-A3B, Sparsity=50%2025.0500.64-80.05
SEER (G)Base Model=Qwen3-30B-A3B, Sparsity=50%2025.0500.34-80.35
Act-LogitsBase Model=Qwen3-30B-A3B, Sparsity=50%2025.0500.82-79.87