Mathematical Reasoning on Math500 (Accuracy, Pass@8)
97.4AccuracyQwen3-8B (Thinking)
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Qwen3-8B (Thinking)Model=Qwen3-8B, Variant=Thinking2026.05 | 97.4 | 100 | |
| Qwen3-1.7B (Thinking)Model=Qwen3-1.7B, Variant=Thinking2026.05 | 88.9 | 100 | |
| Qwen3-1.7B + Guided (ρ¯ ≈ 0.16)Model=Qwen3-1.7B, Intervention Rate (rho)=≈ 0.162026.05 | 81.5 | 100 | |
| Qwen3-0.6B + Guided (ρ¯ ≈ 0.13)Model=Qwen3-0.6B, Intervention Rate (rho)=≈ 0.132026.05 | 73.9 | 99 | |
| Qwen3-0.6B (Thinking)Model=Qwen3-0.6B, Variant=Thinking2026.05 | 72.1 | 90 | |
| Qwen3-1.7B + Guided (ρ¯ ≈ 0.06)Model=Qwen3-1.7B, Intervention Rate (rho)=≈ 0.062026.05 | 66.8 | 95 | |
| Qwen3-1.7B-BaseModel=Qwen3-1.7B, Variant=Base2026.05 | 54.2 | 82 | |
| Qwen3-0.6B + Guided (ρ¯ ≈ 0.04)Model=Qwen3-0.6B, Intervention Rate (rho)=≈ 0.042026.05 | 44.6 | 89 | |
| Qwen3-0.6B-BaseModel=Qwen3-0.6B, Variant=Base2026.05 | 27.6 | 67 |