List your company

Mathematical Reasoning on College Olymp.

36.29Pass@1OpenSIR
28.479630.507332.53534.5627Nov 1, 2025

Evaluation results

MethodLinks
OpenSIRBackbone=Llama-3.1-8B-Instruct2025.1136.29
OpenSIRBackbone=Llama-3.2-3B-Instruct2025.1135.42
GRPOgsm8kBackbone=Llama-3.1-8B-Instruct2025.1135.03
GRPOmathBackbone=Llama-3.1-8B-Instruct2025.1134.93
Absolute ZeroBackbone=Llama-3.1-8B-Instruct2025.1134.39
BaseBackbone=Llama-3.1-8B-Instruct2025.1134.1
GRPOgsm8kBackbone=Llama-3.2-3B-Instruct2025.1133.33
GRPOmathBackbone=Llama-3.2-3B-Instruct2025.1132.95
R-ZeroBackbone=Llama-3.1-8B-Instruct2025.1132.93
R-ZeroBackbone=Llama-3.2-3B-Instruct2025.1132.72
Absolute ZeroBackbone=Llama-3.2-3B-Instruct2025.1131.93
BaseBackbone=Llama-3.2-3B-Instruct2025.1128.78