List your company

Mathematical Reasoning on RealMath (200 held-out questions)

85.1AccuracyMAGE
30.39644.59858.873.002May 11, 2026

Evaluation results

MethodLinks
MAGEBackbone=Qwen3-8B, Judge=Claude Sonnet 4.6, MAGE Executor Model=Sonnet2026.0585.131.6
MAGEBackbone=Qwen3-8B, Judge=Claude Sonnet 4.6, MAGE Executor Model=Opus2026.0582.831.6
0-shot CoTBackbone=Qwen3-8B, Judge=Claude Sonnet 4.6, Prompting Scaffold=0-shot CoT2026.0553.5—
ReActBackbone=Qwen3-8B, Judge=Claude Sonnet 4.6, Prompting Scaffold=ReAct2026.0543.5—
ReflexionBackbone=Qwen3-8B, Judge=Claude Sonnet 4.6, Prompting Scaffold=Reflexion2026.0534.5—
SC10Backbone=Qwen3-8B, Judge=Claude Sonnet 4.6, Prompting Scaffold=SC10, Samples=102026.0533.5—
8-shotBackbone=Qwen3-8B, Judge=Claude Sonnet 4.6, Prompting Scaffold=8-shot2026.0532.5—