| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Free-form Generation | Dragonball | R-L Score0.372 | 12 | |
| Multi-hop Question Answering | Dragonball DragBalance (test) | Recall100 | 11 | |
| Multi-hop Question Answering | Dragonball DragSingleZh (test) | Recall100 | 10 | |
| Information Retrieval | Dragonball | IE Score40.87 | 8 | |
| Question Answering | Dragonball | ROUGE-L38.9 | 5 |