List your company

Mathematical Reasoning on AIME25 (avg@32)

24.38Avg@32Pion
9.528813.384417.2421.0956May 12, 2026

Evaluation results

MethodLinks
PionBase Model=DeepSeek-R1-Distill-Qwen-1.5B2026.0524.38
AdamWBase Model=DeepSeek-R1-Distill-Qwen-1.5B2026.0523.94
MuonBase Model=DeepSeek-R1-Distill-Qwen-1.5B2026.0523.33
PionBase Model=Qwen3-1.7B2026.0521.98
AdamWBase Model=Qwen3-1.7B2026.0520.94
BaseBase Model=DeepSeek-R1-Distill-Qwen-1.5B2026.0520.83
MuonBase Model=Qwen3-1.7B2026.0519.27
BaseBase Model=Qwen3-1.7B2026.0510.1