List your company

Mathematical Reasoning on AIME 2025 (Accuracy, Avg Tokens)

86.67AccuracyGemma-4 31B
13.8732.7751.6770.57Jun 2, 2026

Evaluation results

MethodLinks
Gemma-4 31BEvaluation Mode=Reasoning+Rotate2Think2026.0686.674,240
Gemma-4 31BEvaluation Mode=Reasoning2026.0683.337,170
Qwen3 4BEvaluation Mode=Reasoning+Rotate2Think2026.068020,894
Phi-4 14BEvaluation Mode=Reasoning+Rotate2Think2026.0676.6717,108
Phi-4 14BEvaluation Mode=Reasoning2026.0673.3315,053
Gemma-4 31BEvaluation Mode=Base+Rotate2Think2026.0673.334,110
Qwen3 4BEvaluation Mode=Reasoning2026.067020,944
Gemma-4 31BEvaluation Mode=Base2026.06702,927
Gemma-4 E4BEvaluation Mode=Reasoning+Rotate2Think2026.06505,891
Qwen3 4BEvaluation Mode=Base+Rotate2Think2026.0646.678,299
Qwen3 4BEvaluation Mode=Base2026.0643.3310,444
Gemma-4 E4BEvaluation Mode=Base+Rotate2Think2026.06405,408
Gemma-4 E4BEvaluation Mode=Reasoning2026.06406,968
Gemma-4 E4BEvaluation Mode=Base2026.0636.673,438
Phi-4 14BEvaluation Mode=Base+Rotate2Think2026.06201,639
Phi-4 14BEvaluation Mode=Base2026.0616.671,672