Entailment Tree Generation on EntailmentBank Task 3 (Full Unseen)
47.1Leaves F1SEER
Evaluation Results
| Method | Links | |||||||
|---|---|---|---|---|---|---|---|---|
| SEERBackbone=T5-large2024.01 | 47.1 | 13.8 | 17.4 | 12.9 | 45.1 | 18.8 | 12.9 | |
| GPT4-ReActModel Version=gpt-4-1106-preview, Reasoning Mode=ReAct2024.01 | 45.8 | 12.9 | 14.1 | 10.5 | 43.5 | 21.5 | 10.5 | |
| IRGRBackbone=T5-large2024.01 | 45.6 | 11.8 | 16.1 | 11.4 | 38.8 | 20.9 | 11.5 | |
| GPT4-CoTModel Version=gpt-4-1106-preview, Reasoning Mode=Chain of Thought2024.01 | 44.1 | 12.1 | 15.4 | 10.8 | 43.1 | 20.6 | 10.8 | |
| FAMEBackbone=T5-large2024.01 | 43.4 | 13.8 | 16.6 | 12.4 | 40.6 | 19.9 | 11.9 | |
| GPT4-ToTModel Version=gpt-4-1106-preview, Reasoning Mode=Tree of Thought2024.01 | 43.3 | 12 | 15.8 | 11 | 43.9 | 20 | 11 | |
| NLProofSBackbone=T5-large2024.01 | 43.2 | 8.2 | 11.2 | 6.9 | 42.9 | 17.3 | 6.9 | |
| RLETBackbone=DeBERTa-large2024.01 | 38.3 | 9.1 | 11.5 | 7.1 | 34.2 | 12.1 | 6.9 | |
| EntailmentWriterBackbone=T5-large2024.01 | 35.7 | 2.9 | 6.1 | 2.4 | 33.4 | 7.7 | 2.4 | |
| METGENBackbone=T5-large2024.01 | 34.8 | 8.7 | 9.8 | 8.6 | 36.7 | 20.4 | 8.6 |