List your company

Mathematical Problem Solving on AIME 2024 (test)

28.9AccuracyBudgetThinker
14.23618.04321.8525.657Aug 24, 2025

Evaluation results

MethodLinks
BudgetThinkerGeneration Budget (B)=10000, Model Scale=1.5B2025.0828.9
OriginalGeneration Budget (B)=10000, Model Scale=1.5B2025.0828.88
ThinkPruneGeneration Budget (B)=10000, Model Scale=1.5B2025.0827.8
OriginalGeneration Budget (B)=6000, Model Scale=1.5B2025.0825.71
BudgetThinkerGeneration Budget (B)=6000, Model Scale=1.5B2025.0825.63
ThinkPruneGeneration Budget (B)=6000, Model Scale=1.5B2025.0825.22
BudgetThinkerGeneration Budget (B)=4000, Model Scale=1.5B2025.0823.8
OriginalGeneration Budget (B)=4000, Model Scale=1.5B2025.0823.28
ThinkPruneGeneration Budget (B)=4000, Model Scale=1.5B2025.0822.78
BudgetThinkerGeneration Budget (B)=2000, Model Scale=1.5B2025.0816.25
ThinkPruneGeneration Budget (B)=2000, Model Scale=1.5B2025.0815.22
OriginalGeneration Budget (B)=2000, Model Scale=1.5B2025.0814.8