Cross-domain
Benchmarks
| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Comparative empirical forecasting | Cross-Domain CD (test) | Overall Consistency Rate82.99 | 10 | |
| Zero-shot forecasting on multivariate time series | Cross-Domain Short Forecast All | MAE0.592 | 10 | |
| mol2text (Molecule captioning) | Cross-domain | BLEU0.625 | 8 | |
| text2mol (Text-conditional de novo generation) | Cross-domain | BLEU Score85.3 | 8 | |
| Video Moment Retrieval | T → C cross-domain (Source: T, Target: C) (target domain evaluation) | R@1 (IoU=0.3)48.69 | 7 | |
| Video Moment Retrieval | Cross-domain C → A (target evaluation) | R@1 (IoU=0.3)70.25 | 7 | |
| Video Moment Retrieval | A → C cross-domain (target domain evaluation) | R@1 (IoU=0.3)71.58 | 7 | |
| Trajectory Transfer | Cross-Domain 3D-FRONT to ARKitScenes | Trajectory AED1.464 | 7 | |
| 5-way Classification | Cross-domain 5-shot | Accuracy74.88 | 7 | |
| Zero-shot forecasting on multivariate time series | Cross-Domain Generative Micro Avg. | MAE0.638 | 5 | |
| Zero-shot forecasting on multivariate time series | Cross-Domain Long Simulate All | MAE0.738 | 5 | |
| Zero-shot forecasting on multivariate time series | Cross-Domain Long Forecast All | MAE0.72 | 5 | |
| AI-Generated Text Detection | Cross-domain (test) | OA93 | 4 | |
| Emotion Recognition | Cross-domain (test) | Accuracy61.4 | 3 | |
| Constituency parsing | Cross-domain (Bio, Dialogue, Forum, Law, Literature, Review) (test) | Accuracy (Bio)86.61 | 3 | |
| Threat Modeling | Cross-domain | Metric- | 0 |
Showing 16 of 16 rows