Mathematical Reasoning on Mathematical Reasoning Benchmarks Average
78.1AccuracyQwen3-8B (Thinking)
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| Qwen3-8B (Thinking)Model=Qwen3-8B, Variant=Thinking2026.05 | 78.1 | 87.3 | — | |
| Qwen3-1.7B (Thinking)Model=Qwen3-1.7B, Variant=Thinking2026.05 | 64.1 | 80.3 | — | |
| Qwen3-1.7B + Guided (ρ¯ ≈ 0.16)Model=Qwen3-1.7B, Intervention Rate (rho)=≈ 0.162026.05 | 62.1 | 83.8 | 112 | |
| Qwen3-0.6B + Guided (ρ¯ ≈ 0.13)Model=Qwen3-0.6B, Intervention Rate (rho)=≈ 0.132026.05 | 52.4 | 80 | 157 | |
| Qwen3-1.7B + Guided (ρ¯ ≈ 0.06)Model=Qwen3-1.7B, Intervention Rate (rho)=≈ 0.062026.05 | 43.6 | 73.5 | 76 | |
| Qwen3-0.6B (Thinking)Model=Qwen3-0.6B, Variant=Thinking2026.05 | 43.4 | 64.1 | — | |
| Qwen3-1.7B-BaseModel=Qwen3-1.7B, Variant=Base2026.05 | 30.8 | 51.8 | — | |
| Qwen3-0.6B + Guided (ρ¯ ≈ 0.04)Model=Qwen3-0.6B, Intervention Rate (rho)=≈ 0.042026.05 | 29.1 | 61.4 | 91 | |
| Qwen3-0.6B-BaseModel=Qwen3-0.6B, Variant=Base2026.05 | 13 | 36 | — |