List your company
ResearchFormerly WizWand

Mathematical Reasoning on AMC 23 (Response Length (Tokens))

6,768Response Length (Tokens)DeepSeek-R1-Distill
701.682,276.593,851.55,426.41Jun 16, 2026
Updated 4mo ago

Evaluation results

MethodLinks
DeepSeek-R1-DistillModel Scale=1.5B, Base Model=Qwen2.5-Math2026.066,768
DeepSeek-R1-DistillModel Scale=7B, Base Model=Qwen2.5-Math2026.065,466
AdaCoTModel Scale=7B, Base Model=Qwen2.5-Math2026.063,115
S-GRPOModel Scale=1.5B, Base Model=Qwen2.5-Math2026.063,081
AdaCoTModel Scale=1.5B, Base Model=Qwen2.5-Math2026.062,936
AdaptThinkModel Scale=7B, Base Model=Qwen2.5-Math2026.062,834
AdaptThinkModel Scale=1.5B, Base Model=Qwen2.5-Math2026.062,740
LHRMsModel Scale=1.5B, Base Model=Qwen2.5-Math2026.062,477
S-GRPOModel Scale=7B, Base Model=Qwen2.5-Math2026.062,247
SuCoModel Scale=1.5B, Base Model=Qwen2.5-Math2026.061,687
LHRMsModel Scale=7B, Base Model=Qwen2.5-Math2026.061,525
SuCoModel Scale=7B, Base Model=Qwen2.5-Math2026.06935