| Dataset Name | SOTA Method | Metric | Trend | ||
|---|---|---|---|---|---|
| WritingBench | EvoRubric | Score75.76 | 20 | 2d ago | |
| ArenaHard | Accuracy50 | 17 | 1d ago | ||
| DeepresearchBench | Qwen3-8B + R2-Write-SFT + RLp | Overall Score46.93 | 11 | 1mo ago | |
| HelloBench | Qwen3-8B + R2-Write-SFT + RLp | Average Score82 | 11 | 1mo ago |