Factual Consistency Evaluation on SE
37.4Kendall's TauALIGNSCORE-large
Evaluation Results
| Method | Links | |
|---|---|---|
| ALIGNSCORE-largeType=Ours2023.05 | 37.4 | |
| UniEvalType=Misc2023.05 | 35.3 | |
| DAEType=NLI2023.05 | 34.8 | |
| ALIGNSCORE-baseType=Ours2023.05 | 34.7 | |
| QAFactEvalType=QA2023.05 | 34 | |
| CTCType=Misc2023.05 | 33.2 | |
| SummaC-CONVType=NLI2023.05 | 33.1 | |
| FactCCType=Misc2023.05 | 32.2 | |
| BARTScoreType=Misc2023.05 | 31 | |
| ROUGE-LType=Similarity Matching2023.05 | 30.6 | |
| SummaC-ZSType=NLI2023.05 | 30.4 | |
| ROUGE-1Type=Similarity Matching2023.05 | 30.3 | |
| ROUGE-2Type=Similarity Matching2023.05 | 30 | |
| BLEUType=Similarity Matching2023.05 | 27.5 | |
| BERTScoreType=Similarity Matching2023.05 | 24.9 | |
| SimCSEType=Similarity Matching2023.05 | 20.9 | |
| QuestEvalType=QA2023.05 | 20.8 | |
| NER-OverlapType=Similarity Matching2023.05 | 20.6 | |
| BLEURTType=Regression2023.05 | 18.6 | |
| BLANCType=Misc2023.05 | 14.9 | |
| FEQAType=QA2023.05 | 0.2 | |
| MNLIType=NLI2023.05 | -5.2 |