| Dataset Name | SOTA Method | Metric | Trend | ||
|---|---|---|---|---|---|
| LongBench | HotpotQA Acc57.67 | 45 | 15d ago | ||
| MDQA | SE | Recall@172.6 | 16 | 3mo ago | |
| MultidocQA | MemSearch-o1 | HotpotQA Accuracy67.78 | 14 | 1mo ago | |
| Chinese Multi-doc QA (test) | ChatGLM3-6B-32K | Rouge-L44.8 | 10 | 3mo ago | |
| HotpotQA | Accuracy45 | 9 | 19d ago | ||
| Loong | Average Score68 | 8 | 13d ago | ||
| FRAMES (test) | RADAR | Hypervolume0.8762 | 4 | 2mo ago |