List your company

Mathematical Problem Solving on AIME (Success Rate)

24.2AIME Success RateFreshPER
16.50418.50220.522.498Apr 18, 2026

Evaluation results

MethodLinks
FreshPERType=LLM2026.0424.2
On-PolicyType=LLM2026.0420.5
Standard PERType=LLM2026.0416.8