List your company

Mathematical Reasoning on AIME 2024 (Mean@16, Acc)

79.17Mean@16 AccuracyME-ICPO
18.27834.086549.89565.7035Mar 2, 2026

Evaluation results

MethodLinks
ME-ICPOModel=Gemini-2.5-Pro2026.0379.1780
ME-ICPOModel=Gemini-2.5-Flash2026.0376.4676.47
BaseModel=Gemini-2.5-Pro2026.0358.5456.6
ME-ICPOModel=Qwen3-4B-Instruct2026.0357.7157.67
BaseModel=Gemini-2.5-Flash2026.0335.2135.42
BaseModel=Qwen3-4B-Instruct2026.0320.6220.59