List your company
ResearchFormerly WizWand

Math Reasoning on AIME 2024 (Acc, Tok, AES)

0.627AccuracyExOPD
0.198520.309760.4210.53224Feb 10, 2026Feb 11, 2026Feb 12, 2026
Updated 8mo ago

Evaluation results

MethodLinks
ExOPDDistillation Setting=Single-Teacher Distillation2026.020.627——
ExOPDDistillation Setting=Multi-Teacher Distillation2026.020.61——
OPDDistillation Setting=Single-Teacher Distillation2026.020.607——
OPDDistillation Setting=Multi-Teacher Distillation2026.020.606——
ExPODistillation Setting=Single-Teacher Distillation2026.020.587——
TLMREModel Size=7B, Learning Paradigm=outcome-supervised RL2026.020.5857,4620.61
SFTDistillation Setting=Multi-Teacher Distillation2026.020.585——
Laser-DModel Size=7B, Learning Paradigm=outcome-supervised RL2026.020.5835,3790.79
TeacherDistillation Setting=Baseline (Teacher)2026.020.58——
DIETModel Size=7B, Learning Paradigm=outcome-supervised RL2026.020.57910,1240.32
ExPODistillation Setting=Multi-Teacher Distillation2026.020.575——
ATTNPOModel Size=7B, Learning Paradigm=process-supervised RL2026.020.5725,2140.75
S-GRPOModel Size=7B, Learning Paradigm=process-supervised RL2026.020.567,3770.47
Laser-DEModel Size=7B, Learning Paradigm=outcome-supervised RL2026.020.5584,9690.69
AdaptThinkModel Size=7B, Learning Paradigm=adaptive-mode2026.020.5568,5990.33
AutoThinkModel Size=7B, Learning Paradigm=adaptive-mode2026.020.5488,0510.34
DS-7BModel Size=7B2026.020.52910,4900
ACPOModel Size=7B, Learning Paradigm=outcome-supervised RL2026.020.5284,5200.56
DEPOModel Size=7B, Learning Paradigm=process-supervised RL2026.020.5276,5800.35
VSRM-R++Model Size=7B, Learning Paradigm=process-supervised RL2026.020.5226,7730.29
DECSModel Size=7B, Learning Paradigm=process-supervised RL2026.020.5135,2770.35
ATTNPOModel Size=1.5B, Learning Paradigm=process-supervised RL2026.020.3775,7131.56
TLMREModel Size=1.5B, Learning Paradigm=outcome-supervised RL2026.020.3737,0241.41
Laser-DEModel Size=1.5B, Learning Paradigm=outcome-supervised RL2026.020.355,7891.27
AutoThinkModel Size=1.5B, Learning Paradigm=adaptive-mode2026.020.3469,5140.92
Laser-DModel Size=1.5B, Learning Paradigm=outcome-supervised RL2026.020.3425,7501.19
DIETModel Size=1.5B, Learning Paradigm=outcome-supervised RL2026.020.31810,5780.54
DECSModel Size=1.5B, Learning Paradigm=process-supervised RL2026.020.3125,5500.89
AdaptThinkModel Size=1.5B, Learning Paradigm=adaptive-mode2026.020.316,6700.77
DEPOModel Size=1.5B, Learning Paradigm=process-supervised RL2026.020.3087,7320.66
ACPOModel Size=1.5B, Learning Paradigm=outcome-supervised RL2026.020.36,6700.66
VSRM-R++Model Size=1.5B, Learning Paradigm=process-supervised RL2026.020.2956,9580.59
DS-1.5BModel Size=1.5B2026.020.28112,3390
ThinkPruneModel Size=1.5B, Learning Paradigm=outcome-supervised RL2026.020.2715,6310.37
LC-R1Model Size=1.5B, Learning Paradigm=process-supervised RL2026.020.2367,122-0.37
LC-R1Model Size=7B, Learning Paradigm=process-supervised RL2026.020.2366,904-2.42
StudentDistillation Setting=Baseline (Student)2026.020.215——