Code Generation on HumanEval Online Collection
82.4pass@1EXG-Reflexion
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| EXG-ReflexionBackbone=Qwen3-Coder-Flash2026.05 | 82.4 | 88.5 | 1.35 | |
| ExpeLBackbone=Qwen3-Coder-Flash2026.05 | 57.3 | 77.1 | 1.85 |
Loading the SOTA2 catalog…
| Method | Links | |||
|---|---|---|---|---|
| EXG-ReflexionBackbone=Qwen3-Coder-Flash2026.05 | 82.4 | 88.5 | 1.35 | |
| ExpeLBackbone=Qwen3-Coder-Flash2026.05 | 57.3 | 77.1 | 1.85 |