Factuality Evaluation on QAGS CNN
0.735Pearson CorrelationBARTSCORE
Evaluation Results
| Method | Links | |
|---|---|---|
| BARTSCOREFine-tuning task=CNN2021.06 | 0.735 | |
| BARTSCOREFine-tuning task=CNN, Prompting=true2021.06 | 0.719 | |
| BARTSCOREFine-tuning task=CNN + Para2021.06 | 0.68 | |
| BARTSCORESetup=Vanilla2021.06 | 0.661 | |
| BERTScore2021.06 | 0.576 | |
| QAGS2021.06 | 0.545 | |
| PRISM2021.06 | 0.479 | |
| ROUGE-22021.06 | 0.459 | |
| MoverScore2021.06 | 0.414 | |
| ROUGE-L2021.06 | 0.357 | |
| ROUGE-12021.06 | 0.338 |