List your company

Mathematical Reasoning on AIME 2024 (ACC, Avg Tokens)

81.48ACC (%)Qwen3-8B
55.677662.376369.07575.7737Aug 5, 2025

Evaluation results

MethodLinks
Qwen3-8Bstrategy=step-entropy pruning, pruning_ratio=80%2025.0881.4811,533.57
Qwen3-8Bstrategy=Full COT2025.0879.3120,936.57
QwQ-32Bstrategy=Full COT2025.0879.1321,243.93
QwQ-32Bstrategy=step-entropy pruning, pruning_ratio=80%2025.0874.079,544.13
DeepSeek-R1-14Bstrategy=Full COT2025.0865.5215,414.83
DeepSeek-R1-7Bstrategy=Full COT2025.0863.3315,843.43
DeepSeek-R1-14Bstrategy=step-entropy pruning, pruning_ratio=80%2025.0858.628,705.57
DeepSeek-R1-7Bstrategy=step-entropy pruning, pruning_ratio=80%2025.0856.6710,092.8