Coding on LiveCodeBench Jan 25 – May 25
42AccuracyQwen3-4B-REVES
Evaluation results
| Method | Links | |
|---|---|---|
| Qwen3-4B-REVESBudget=TC-322026.06 | 42 | |
| Qwen3-4B-Multi-turnBudget=TC-322026.06 | 40.9 | |
| Qwen3-4B-PAGBudget=TC-322026.06 | 37.1 | |
| Qwen3-4B-RLBudget=TC-322026.06 | 35.4 | |
| Qwen-2.5-7B-REVESBudget=TC-322026.06 | 30 | |
| Qwen3-4B-REVESBudget=OneShot2026.06 | 29.7 | |
| Qwen3-4B-RLBudget=OneShot2026.06 | 29.1 | |
| Qwen-2.5-7B-Multi-turnBudget=TC-322026.06 | 27.5 | |
| Qwen3-4B-PAGBudget=OneShot2026.06 | 26.7 | |
| Qwen-2.5-7B-PAGBudget=TC-322026.06 | 26.4 | |
| Qwen3-4B-Multi-turnBudget=OneShot2026.06 | 25.3 | |
| Qwen3-4B (base)Budget=OneShot2026.06 | 24.7 | |
| Qwen3-4B (base)Budget=TC-322026.06 | 24.7 | |
| Qwen-2.5-7B-RLBudget=TC-322026.06 | 24.2 | |
| Qwen-2.5-3B-REVESBudget=TC-322026.06 | 21.6 | |
| Qwen-2.5-7B-Multi-turnBudget=OneShot2026.06 | 20.7 | |
| Qwen-2.5-7B-REVESBudget=OneShot2026.06 | 20.5 | |
| Qwen-2.5-7B-RLBudget=OneShot2026.06 | 20 | |
| Qwen-2.5-7B-PAGBudget=OneShot2026.06 | 19.8 | |
| Qwen-2.5-3B-Multi-turnBudget=TC-322026.06 | 19.2 | |
| Qwen-2.5-7BBudget=TC-322026.06 | 18.9 | |
| Qwen-2.5-3B-PAGBudget=TC-322026.06 | 18.5 | |
| Qwen-2.5-3BBudget=TC-322026.06 | 18.3 | |
| Qwen-2.5-3B-RLBudget=TC-322026.06 | 17.8 | |
| Qwen-2.5-3B-REVESBudget=OneShot2026.06 | 14.3 | |
| Qwen-2.5-3B-RLBudget=OneShot2026.06 | 14.1 | |
| Qwen-2.5-3B-PAGBudget=OneShot2026.06 | 13.4 | |
| Qwen-2.5-3B-Multi-turnBudget=OneShot2026.06 | 13.2 | |
| Qwen-2.5-3BBudget=OneShot2026.06 | 7.9 | |
| Qwen-2.5-7BBudget=OneShot2026.06 | 6 |