List your company

Mathematical Reasoning on AIME 25 (Accuracy and Token Count)

24.38AIME 25 AccuracyMUR
9.643213.469117.29521.1209Jul 20, 2025

Evaluation results

MethodLinks
MURBackbone=Qwen3-8B, Search Strategy=phi-Decoding2025.0724.3822,296
MURBackbone=Qwen3-4B, Search Strategy=LLM as a critic2025.0720.213,113
MURBackbone=Qwen3-1.7B, Search Strategy=LLM as a critic2025.0710.214,011