List your company
ResearchFormerly WizWand

Mathematical Reasoning on AIME 22-24 (Pass@1, #Tokens)

45.1Pass@1Dynasor
10.88419.76728.6537.533Dec 31, 2025
Updated 8mo ago

Evaluation results

MethodLinks
DynasorModel=R1-7B2025.1245.110,314
VanillaModel=R1-7B2025.1244.413,709
TIPModel=R1-7B2025.1244.411,112
CRESTModel=R1-7B2025.1244.49,488
SEALModel=R1-7B2025.1242.2212,448
Soft-ThinkingModel=R1-7B2025.1235.612,551
TIPModel=R1-1.5B2025.1224.414,157
DynasorModel=R1-1.5B2025.1224.1215,337
SEALModel=R1-1.5B2025.1222.214,207
CRESTModel=R1-1.5B2025.122013,388
VanillaModel=R1-1.5B2025.1217.817,034
Soft-ThinkingModel=R1-1.5B2025.1212.218,418