List your company

Mathematical Reasoning on Math500 (Accuracy, Pass@8)

97.4AccuracyQwen3-8B (Thinking)
24.80843.65462.581.346May 16, 2026

Evaluation results

MethodLinks
Qwen3-8B (Thinking)Model=Qwen3-8B, Variant=Thinking2026.0597.4100
Qwen3-1.7B (Thinking)Model=Qwen3-1.7B, Variant=Thinking2026.0588.9100
Qwen3-1.7B + Guided (ρ¯ ≈ 0.16)Model=Qwen3-1.7B, Intervention Rate (rho)=≈ 0.162026.0581.5100
Qwen3-0.6B + Guided (ρ¯ ≈ 0.13)Model=Qwen3-0.6B, Intervention Rate (rho)=≈ 0.132026.0573.999
Qwen3-0.6B (Thinking)Model=Qwen3-0.6B, Variant=Thinking2026.0572.190
Qwen3-1.7B + Guided (ρ¯ ≈ 0.06)Model=Qwen3-1.7B, Intervention Rate (rho)=≈ 0.062026.0566.895
Qwen3-1.7B-BaseModel=Qwen3-1.7B, Variant=Base2026.0554.282
Qwen3-0.6B + Guided (ρ¯ ≈ 0.04)Model=Qwen3-0.6B, Intervention Rate (rho)=≈ 0.042026.0544.689
Qwen3-0.6B-BaseModel=Qwen3-0.6B, Variant=Base2026.0527.667