List your company

Mathematical Reasoning on MATH, Minerva, Olympiad, AIME24, AIME25, AMC23

70.65MATH ScoreRASFT
-1.255617.412236.0854.7478Jun 5, 2026

Evaluation results

MethodLinks
RASFTBackbone=Qwen2.5-Math-7B2026.0670.6529.9732.998.765.8445.7832.33
ProFiTBackbone=Qwen2.5-Math-7B2026.0668.4528.6731.417.084.9944.0630.78
DFTBackbone=Qwen2.5-Math-7B2026.0668.2832.9632.068.974.845.3132.06
RASFTBackbone=Qwen2.5-Math-1.5B2026.0666.3923.5927.17.923.9637.3427.72
ASFTBackbone=Qwen2.5-Math-7B2026.0665.7624.1429.258.975.8343.2829.54
ASFTBackbone=Qwen2.5-Math-1.5B2026.0660.8920.1824.114.374.3736.0925
DFTBackbone=Qwen2.5-Math-1.5B2026.0660.5820.4122.424.793.3233.5924.19
ProFiTBackbone=Qwen2.5-Math-1.5B2026.0658.422.7922.843.742.2832.9723.84
SFTBackbone=Qwen2.5-Math-7B2026.0651.5217.4916.381.881.0427.9719.38
SFTBackbone=Qwen2.5-Math-1.5B2026.0640.6911.2710.281.660.4116.5613.48
BaseBackbone=Qwen2.5-Math-7B2026.0632.397.929.017.291.8719.2212.95
BaseBackbone=Qwen2.5-Math-1.5B2026.0623.215.8213.241.871.4618.2810.65
ProFiTBackbone=Llama-3.2-3B2026.068.513.072.260.210.214.383.11
RASFTBackbone=Llama-3.2-3B2026.068.263.991.980.210.214.843.25
DFTBackbone=Llama-3.2-3B2026.067.713.072.030.210.212.192.57
SFTBackbone=Llama-3.2-3B2026.064.861.581.24001.091.46
ASFTBackbone=Llama-3.2-3B2026.064.351.991.420.210.212.661.81
BaseBackbone=Llama-3.2-3B2026.061.511.150.8600.411.560.92