List your company

Mathematical Reasoning on AIME24 (Accuracy, Performance Gain %)

75AccuracyQwen3-32B
38.648.0557.566.95Jun 4, 2026

Evaluation results

MethodLinks
Qwen3-32BSetting=Recur2026.06755.81
Qwen3-14BSetting=Recur2026.0673.3311.43
Qwen3-8BSetting=Recur2026.0671.6612.57
Qwen3-32BSetting=Base2026.0670.83—
Qwen3-4BSetting=Recur2026.067016.67
LLaMA-70BSetting=Recur2026.0666.675.25
Qwen3-14BSetting=Base2026.0665.83—
Qwen3-8BSetting=Base2026.0663.67—
LLaMA-70BSetting=Base2026.0663.33—
Qwen3-4BSetting=Base2026.0660—
Qwen3-1.7BSetting=Recur2026.0648.3320.83
LLaMA-8BSetting=Recur2026.0647.518.75
Qwen3-1.7BSetting=Base2026.0640—
LLaMA-8BSetting=Base2026.0640—