| Dataset Name | SOTA Method | Metric | Trend | ||
|---|---|---|---|---|---|
| MMLU | Accuracy89.5 | 27 | 26d ago | ||
| Humanity's Last Exam (HLE) text-only | sHLE Score36.6 | 11 | 4mo ago | ||
| CEval | Accuracy90.91 | 4 | 3mo ago | ||
| General Tasks Suite BBH, MMLU, CMMLU, C-Eval | Qwen3-Inst | BBH59.48 | 4 | 4mo ago | |
| MMLU-Pro | StructU+Self-ConsU | Sel-AUC94.4 | 2 | 1mo ago |