List your company

Mathematical Reasoning on AMC 23 (Pass@16, Mean@16)

55.57Mean@16ADS
21.125230.067639.0147.9524Jun 21, 2026

Evaluation results

MethodLinks
ADSBackbone=Olmo3-7B-SFT2026.0655.5795
ADSBackbone=Qwen3-4B-Base2026.0654.8989.17
GRPOBackbone=Qwen3-4B-Base2026.0653.6589.17
DOTS+RRBackbone=Qwen3-4B-Base2026.0652.9189.17
DOTS+RRBackbone=Olmo3-7B-SFT2026.0649.2285.83
ADSBackbone=Qwen2.5-Math-1.5B2026.0645.8987.5
DOTS+RRBackbone=Qwen2.5-Math-1.5B2026.0644.8983.33
GRPOBackbone=Olmo3-7B-SFT2026.0640.4784.17
GRPOBackbone=Qwen2.5-Math-1.5B2026.0635.6879.17
BaseBackbone=Olmo3-7B-SFT2026.0634.980.83
BaseBackbone=Qwen3-4B-Base2026.0623.778.33
BaseBackbone=Qwen2.5-Math-1.5B2026.0622.4580.83