| Benchmark | Model | Rank | Results |
|---|---|---|---|
| code-generation-on-swe-bench-pro | Self-HPO-Qwen3.5-35B-A3B | #6 | Resolved: 65.2 |
| code-generation-on-swe-bench-multilingual | Self-HPO-Qwen3.5-35B-A3B | #30 | Accuracy: 59.7 |
| code-generation-on-swe-bench-verified | Self-HPO-Qwen3.5-35B-A3B | #45 | Accuracy: 69.5 |