List your company
ResearchFormerly WizWand

Mathematical Reasoning on Average (MATH-OAI, Minerva Math, OlympiadBench, AIME24, AMC23)

32.69Average Score @16PriFT-mass
0.77249.058717.34525.6313Jun 8, 2026
Updated 4mo ago

Evaluation results

MethodLinks
PriFT-massModel=Qwen2.5-Math (1.5B)2026.0632.6964.52
TALRModel=Qwen2.5-Math (1.5B)2026.0631.9557.41
DFTModel=Qwen2.5-Math (1.5B)2026.0631.4154.23
PriFT-probModel=Qwen2.5-Math (1.5B)2026.0631.3863.71
IDFTModel=Qwen2.5-Math (1.5B)2026.0630.9854.51
ASFTModel=Qwen2.5-Math (1.5B)2026.0627.8159.22
PriFT-massModel=Qwen2.5-Instruct (1.5B)2026.0623.7551.01
PriFT-probModel=Qwen2.5-Instruct (1.5B)2026.0621.7452.52
DFTModel=DeepSeekMath (7B)2026.0620.5745.31
PriFT-massModel=DeepSeekMath (7B)2026.0620.1148.22
PriFT-probModel=DeepSeekMath (7B)2026.0619.5949.34
DFTModel=Qwen2.5-Instruct (1.5B)2026.0618.9643.78
PretrainedModel=Qwen2.5-Instruct (1.5B)2026.0618.549.56
DFTModel=Qwen2.5 (1.5B)2026.0618.4844.33
PriFT-massModel=Qwen2.5 (1.5B)2026.0617.9647.59
SFTModel=Qwen2.5-Math (1.5B)2026.0617.8450.28
EAFTModel=Qwen2.5-Math (1.5B)2026.0617.5850.22
PretrainedModel=DeepSeekMath (7B)2026.0616.1747.45
PriFT-probModel=Qwen2.5 (1.5B)2026.0615.8846.13
OriginalModel=Qwen2.5-Math (1.5B)2026.0615.157.25
SFTModel=DeepSeekMath (7B)2026.0612.8242.99
SFTModel=Qwen2.5-Instruct (1.5B)2026.069.0839.49
SFTModel=Qwen2.5 (1.5B)2026.068.7735.9
PretrainedModel=Qwen2.5 (1.5B)2026.06219.31