List your company

Mathematical Reasoning on MATH500 (Average Accuracy, Pass@3)

92.07Average Accuracy (MATH500)GRAIL
41.94254.95667.9780.984Jun 3, 2026

Evaluation results

MethodLinks
GRAILModel=Qwen3-8B, Sampling Temperature=0.6, Generations per problem=32026.0692.0795.4
GRAILModel=Qwen3-4B, Sampling Temperature=0.6, Generations per problem=32026.0691.4793.8
GRPOModel=Qwen3-8B, Sampling Temperature=0.6, Generations per problem=32026.0690.0795
GRPOModel=Qwen3-4B, Sampling Temperature=0.6, Generations per problem=32026.0688.0793.6
GRAILModel=R1-Distill-Llama-8B, Sampling Temperature=0.6, Generations per problem=32026.0682.691
GRPOModel=R1-Distill-Llama-8B, Sampling Temperature=0.6, Generations per problem=32026.068189
GRAILModel=OctoThinker-8B, Sampling Temperature=0.6, Generations per problem=32026.0654.5368.4
GRPOModel=OctoThinker-8B, Sampling Temperature=0.6, Generations per problem=32026.0648.5365.2
GRAILModel=OctoThinker-3B, Sampling Temperature=0.6, Generations per problem=32026.0646.461.4
GRPOModel=OctoThinker-3B, Sampling Temperature=0.6, Generations per problem=32026.0643.8759.6