List your company

Mathematical Reasoning on MATH 500 (pass@1, pass@32)

90.4Pass@1 RateClip-Cov
50.67260.98671.381.614Oct 13, 2025

Evaluation results

MethodLinks
Clip-CovModel Backbone=Qwen3-8B-Base2025.1090.497.8
AERModel Backbone=Qwen3-8B-Base2025.1089.497
Clip-HigherModel Backbone=Qwen3-8B-Base2025.1088.397
Clip-CovModel Backbone=Qwen3-4B-Base2025.1087.895.9
AERModel Backbone=Qwen3-4B-Base2025.1086.796.6
Ent-AdvModel Backbone=Qwen3-8B-Base2025.1083.593.4
GRPOModel Backbone=Qwen3-8B-Base2025.108294.1
Clip-HigherModel Backbone=Qwen3-4B-Base2025.1081.193.5
KL-CovModel Backbone=Qwen3-8B-Base2025.1080.696.7
GRPOModel Backbone=Qwen3-4B-Base2025.1080.593.7
KL-CovModel Backbone=Qwen3-4B-Base2025.1080.395.5
Ent-AdvModel Backbone=Qwen3-4B-Base2025.1079.291.3
BaseModel Backbone=Qwen3-8B-Base2025.1067.294.2
BaseModel Backbone=Qwen3-4B-Base2025.1052.292