List your company

Mathematical Reasoning on AMC 23 (pass@1, pass@32)

75.6Pass@1AER
32.9644.0355.166.17Oct 13, 2025

Evaluation results

MethodLinks
AERModel Backbone=Qwen3-8B-Base2025.1075.694.8
Clip-CovModel Backbone=Qwen3-8B-Base2025.1074.594.1
AERModel Backbone=Qwen3-4B-Base2025.1070.294.8
Clip-HigherModel Backbone=Qwen3-8B-Base2025.1068.592.7
Clip-CovModel Backbone=Qwen3-4B-Base2025.1065.992.4
KL-CovModel Backbone=Qwen3-8B-Base2025.1065.392.1
GRPOModel Backbone=Qwen3-8B-Base2025.1062.888.4
Ent-AdvModel Backbone=Qwen3-8B-Base2025.1062.580.6
GRPOModel Backbone=Qwen3-4B-Base2025.1059.987.8
Clip-HigherModel Backbone=Qwen3-4B-Base2025.1059.689.7
KL-CovModel Backbone=Qwen3-4B-Base2025.1057.284.9
Ent-AdvModel Backbone=Qwen3-4B-Base2025.1049.286.7
BaseModel Backbone=Qwen3-8B-Base2025.104588.7
BaseModel Backbone=Qwen3-4B-Base2025.1034.686.7