List your company

Mathematical Reasoning on AIME 24 (ACC, LEN, delta_LEN)

65.56AccuracyCER
18.177630.478842.7855.0812Oct 2, 2025

Evaluation results

MethodLinks
CERModel=Qwen3-8B2025.1065.569,171.5618.46
w thinkingModel=Qwen3-4B2025.1064.4411,343.57—
w thinkingModel=Qwen3-8B2025.1063.3311,247.68—
CCoTModel=Qwen3-8B2025.1063.339,286.6317.44
CERModel=Qwen3-4B2025.1061.119,215.7718.76
CCoTModel=Qwen3-4B2025.1056.679,491.8716.32
w/o thinkingModel=Qwen3-4B2025.1026.672,132.278.34
CoDModel=Qwen3-4B2025.1023.33,607.8368.19
w/o thinkingModel=Qwen3-8B2025.10202,399.1380.1
CoDModel=Qwen3-8B2025.10203,657.5767.48