List your company

Mathematical Reasoning on BeyondAIME (Turn 1, Turn 2, Δ)

2Turn 1 ScoreUntrained baseline
1.91.9522.05May 23, 2026

Evaluation results

MethodLinks
Untrained baselineBase Model=Qwen3-1.7B2026.05220
Bi-NACBase Model=Qwen3-1.7B2026.05242