List your company
ResearchFormerly WizWand

Mathematical Reasoning on MATH (test) (Accuracy)

84.27MATH AccuracyFAA
31.146844.938458.7372.5216Dec 26, 2025
Updated 8mo ago

Evaluation results

MethodLinks
FAABackbone Model=Deepseek R1-1.5B, Trainable Parameters=1.63M2025.1284.27
FourierFTBackbone Model=Deepseek R1-1.5B, Trainable Parameters=0.15M2025.1284.23
FFBackbone Model=Deepseek R1-1.5B, Trainable Parameters=1.5B2025.1284.21
AdapterHBackbone Model=Deepseek R1-1.5B, Trainable Parameters=1.63M2025.1284.17
LoRABackbone Model=Deepseek R1-1.5B, Trainable Parameters=1.21M2025.1284.16
FAABackbone Model=Qwen2 7B, Trainable Parameters=7.30M2025.1264.18
FourierFTBackbone Model=Qwen2 7B, Trainable Parameters=0.85M2025.1264.12
LoRABackbone Model=Qwen2 7B, Trainable Parameters=5.40M2025.1264.12
FFBackbone Model=Qwen2 7B, Trainable Parameters=7.07B2025.1264.11
AdapterHBackbone Model=Qwen2 7B, Trainable Parameters=7.29M2025.1264.07
FAABackbone Model=LLaMA3 8B, Trainable Parameters=8.73M2025.1246.68
FFBackbone Model=LLaMA3 8B, Trainable Parameters=8.03B2025.1246.61
FourierFTBackbone Model=LLaMA3 8B, Trainable Parameters=0.91M2025.1246.57
LoRABackbone Model=LLaMA3 8B, Trainable Parameters=6.47M2025.1246.53
AdapterHBackbone Model=LLaMA3 8B, Trainable Parameters=8.73M2025.1246.52
FourierFTBackbone Model=LLaMA2 7B, Trainable Parameters=0.82M2025.1233.25
FAABackbone Model=LLaMA2 7B, Trainable Parameters=7.27M2025.1233.24
LoRABackbone Model=LLaMA2 7B, Trainable Parameters=5.37M2025.1233.22
FFBackbone Model=LLaMA2 7B, Trainable Parameters=6.94B2025.1233.21
AdapterHBackbone Model=LLaMA2 7B, Trainable Parameters=7.27M2025.1233.19