List your company

Mathematical Reasoning on AIME 25 (Mean@16, Pass@16)

18.96Mean@16ADS
1.05125.700610.3514.9994Jun 21, 2026

Evaluation results

MethodLinks
ADSBackbone=Olmo3-7B-SFT2026.0618.9645.55
DOTS+RRBackbone=Olmo3-7B-SFT2026.0618.4744.45
ADSBackbone=Qwen3-4B-Base2026.0611.5337.78
GRPOBackbone=Qwen3-4B-Base2026.0611.3237.78
DOTS+RRBackbone=Qwen3-4B-Base2026.069.3130
ADSBackbone=Qwen2.5-Math-1.5B2026.067.0830
DOTS+RRBackbone=Qwen2.5-Math-1.5B2026.066.8833.33
GRPOBackbone=Olmo3-7B-SFT2026.065.5624.44
BaseBackbone=Olmo3-7B-SFT2026.065.4227.78
BaseBackbone=Qwen3-4B-Base2026.064.2327.78
GRPOBackbone=Qwen2.5-Math-1.5B2026.063.4720
BaseBackbone=Qwen2.5-Math-1.5B2026.061.7413.33