Factuality Evaluation on AggreFact-CNN (FTS)
70.3Balanced AccuracySummaC-Cv
Evaluation Results
| Method | Links | |
|---|---|---|
| SummaC-Cvvariant=Convolutional2024.03 | 70.3 | |
| FENICEGPT_claimsExtractor=LLM-based claims2024.03 | 68.2 | |
| ChatGPT-ZSPrompting=Zero-shot2024.03 | 66.2 | |
| TrueTeacher-11BParameters=11B2024.03 | 65.7 | |
| QuestEval2024.03 | 63.7 | |
| SummaC-ZSvariant=Zero-shot2024.03 | 63.3 | |
| QAFactEval2024.03 | 61.6 | |
| FENICET5_claimsExtractor=Knowledge-distilled claim extractor2024.03 | 61.6 | |
| DAENote=Trained on XSumFaith2024.03 | 59.4 | |
| ChatGPT-StarPrompting=Scale-based2024.03 | 55.8 | |
| AlignScore2024.03 | 53.5 | |
| MENLI2024.03 | 51.7 | |
| Random Baseline2024.03 | 50 | |
| ChatGPT-CoTPrompting=Chain-of-Thought2024.03 | 49.7 | |
| ChatGPT-DAPrompting=Direct Assessment2024.03 | 48 |