Factual Consistency Evaluation on XSum-Faithful (XSF)
47Spearman CorrelationDAE
Evaluation Results
| Method | Links | |
|---|---|---|
| DAEType=NLI2023.05 | 47 | |
| ALIGNSCORE-baseType=Ours2023.05 | 43.8 | |
| QuestEvalType=QA2023.05 | 42.1 | |
| BLEURTType=Regression2023.05 | 37 | |
| ROUGE-1Type=Similarity Matching2023.05 | 34.2 | |
| ALIGNSCORE-largeType=Ours2023.05 | 33.3 | |
| QAFactEvalType=QA2023.05 | 31.9 | |
| CTCType=Misc2023.05 | 29.8 | |
| BARTScoreType=Misc2023.05 | 29.8 | |
| SimCSEType=Similarity Matching2023.05 | 29.2 | |
| ROUGE-LType=Similarity Matching2023.05 | 28.9 | |
| ROUGE-2Type=Similarity Matching2023.05 | 26.8 | |
| UniEvalType=Misc2023.05 | 25.3 | |
| NER-OverlapType=Similarity Matching2023.05 | 23.9 | |
| SummaC-CONVType=NLI2023.05 | 21.7 | |
| BLEUType=Similarity Matching2023.05 | 18.2 | |
| BERTScoreType=Similarity Matching2023.05 | 13.4 | |
| BLANCType=Misc2023.05 | 8.4 | |
| MNLIType=NLI2023.05 | 7 | |
| FactCCType=Misc2023.05 | 6.8 | |
| SummaC-ZSType=NLI2023.05 | 5.7 | |
| FEQAType=QA2023.05 | 1.7 |