Factual Consistency Evaluation on FRANK-XSum (FRK-X)
32.1Spearman CorrelationDAE
Evaluation Results
| Method | Links | |
|---|---|---|
| DAEType=NLI2023.05 | 32.1 | |
| ALIGNModel size=large2023.07 | 30 | |
| ALIGNSCORE-largeType=Ours2023.05 | 29 | |
| ALIGNSCORE-baseType=Ours2023.05 | 28 | |
| UniEvalType=Misc2023.05 | 26.7 | |
| UniEval2023.07 | 26.7 | |
| QAFactEvalType=QA2023.05 | 25.5 | |
| QAFactEval2023.07 | 25.5 | |
| ALIGNModel size=base2023.07 | 25.5 | |
| CTCType=Misc2023.05 | 20.4 | |
| CTC2023.07 | 20.4 | |
| BARTScoreType=Misc2023.05 | 20 | |
| BARTScore2023.07 | 20 | |
| QuestEvalType=QA2023.05 | 19.1 | |
| BLEURTType=Regression2023.05 | 13.9 | |
| BLEURT2023.07 | 13.9 | |
| SimCSEType=Similarity Matching2023.05 | 13.3 | |
| SummaC-ZSType=NLI2023.05 | 12.8 | |
| BERTScoreType=Similarity Matching2023.05 | 12.7 | |
| BERTScore2023.07 | 12.7 | |
| MNLIType=NLI2023.05 | 11.7 | |
| NER-OverlapType=Similarity Matching2023.05 | 11.3 | |
| SummaC-CONVType=NLI2023.05 | 11 | |
| ROUGE-LType=Similarity Matching2023.05 | 8.2 | |
| FactCCType=Misc2023.05 | 7.9 | |
| BLANCType=Misc2023.05 | 6.5 | |
| BLEUType=Similarity Matching2023.05 | 6.3 | |
| ROUGE-1Type=Similarity Matching2023.05 | 5.6 | |
| ROUGE-2Type=Similarity Matching2023.05 | 2.8 | |
| FEQAType=QA2023.05 | 1.5 |