List your company

Mathematical Problem Solving on MATH 500 (Accuracy, Refinement Gap)

93.5AccuracyBoN@4
75.19679.94884.789.452Aug 27, 2025

Evaluation results

MethodLinks
BoN@4backbone=GSR-7B, Aggregation Model=RM-R1-7B2025.0893.5—
selfRef@4backbone=GSR-7B2025.0893.40.6
maj@5backbone=GSR-7B2025.0893.1—
maj@4backbone=GSR-7B2025.0892.8—
BoN@5backbone=GSR-7B, Aggregation Model=SkyworkRM-27B2025.0892.7—
BoN@4backbone=GSR-7B, Aggregation Model=SkyworkRM-27B2025.0892.5—
Ref@4backbone=GSR-7B, Aggregation Model=Synthesizer-8B2025.0892—
QwQ-Previewmode=pass@12025.0890.6—
pass@1backbone=GSR-7B2025.0890.6—
BoN@4backbone=GSR-7B, Aggregation Model=RRM-7B2025.0890.6—
o1-minimode=pass@12025.0890—
maj@4backbone=Qwen2.5-7B-Instruct (Base)2025.0879.4—
selfRef@4backbone=Qwen2.5-7B-Instruct (Base)2025.0878.7-0.7
BoN@4backbone=Qwen2.5-7B-Instruct (Base), Aggregation Model=SkyworkRM-27B2025.0878.6—
pass@1backbone=Qwen2.5-7B-Instruct (Base)2025.0875.9—