List your company

Math Reasoning on SuperGPQA SGPQA-1k Math

46.5AccuracyDFT
10.20419.62729.0538.473Oct 31, 2025

Evaluation results

MethodLinks
DFTModel=Qwen3 32B2025.1046.546.5
VCOREModel=Qwen3 32B2025.104545
OriginalModel=Qwen3 32B2025.1043.943.9
RandomModel=Qwen3 32B2025.1040.340.3
SFTModel=Qwen3 32B2025.1039.839.8
iw-SFTModel=Qwen3 32B2025.1039.439.4
DFTModel=Qwen3 8B2025.1037.937.9
VCOREModel=Qwen3 8B2025.1036.836.8
RandomModel=Qwen3 8B2025.1036.236.2
iw-SFTModel=Qwen3 8B2025.1034.634.6
VCOREModel=Qwen3 4B2025.1034.334.3
OriginalModel=Qwen3 8B2025.1034.334.3
SFTModel=Qwen3 8B2025.103434
DFTModel=Qwen3 4B2025.1032.432.4
iw-SFTModel=Qwen3 4B2025.103131
SFTModel=Qwen3 4B2025.1030.530.5
RandomModel=Qwen3 4B2025.1030.530.5
OriginalModel=Qwen3 4B2025.1029.129.1
OriginalModel=LLaMA3.1 8B Instruct2025.1018.618.6
RandomModel=LLaMA3.1 8B Instruct2025.1015.115.1
DFTModel=LLaMA3.1 8B Instruct2025.1013.913.9
VCOREModel=LLaMA3.1 8B Instruct2025.1012.812.8
SFTModel=LLaMA3.1 8B Instruct2025.1012.112.1
iw-SFTModel=LLaMA3.1 8B Instruct2025.1011.611.6