Natural Language Inference on ANLI R1 1.0 (test)
41.1Weighted F1Label-Guided Explanation Selection
Evaluation Results
| Method | Links | |
|---|---|---|
| Label-Guided Explanation SelectionBackbone=RoBERTa, LLM=Llama3, Explanations=Model, Selection Strategy=VariErr Label-Guided2024.12 | 41.1 | |
| VariErr distributionBackbone=RoBERTa2024.12 | 40.2 | |
| MJD EstimatorBackbone=RoBERTa, LLM=Llama3, Explanations=Human2024.12 | 40 | |
| MJD EstimatorBackbone=RoBERTa, LLM=GPT-4o, Explanations=Human2024.12 | 39.3 | |
| Label-Guided Explanation SelectionBackbone=RoBERTa, LLM=GPT-4o, Explanations=Model, Selection Strategy=VariErr Label-Guided2024.12 | 39.3 | |
| ChaosNLI HJDBackbone=RoBERTa2024.12 | 35.7 | |
| MJD EstimatorBackbone=BERT, LLM=GPT-4o, Explanations=Human2024.12 | 35.1 | |
| Label-Free Explanation SelectionBackbone=RoBERTa, LLM=GPT-4o, Explanations=Model2024.12 | 35 | |
| Label-Guided Explanation SelectionBackbone=BERT, LLM=GPT-4o, Explanations=Model, Selection Strategy=VariErr Label-Guided2024.12 | 34.1 | |
| Label-Guided Explanation SelectionBackbone=RoBERTa, LLM=Llama3, Explanations=Model, Selection Strategy=MNLI Label-Guided2024.12 | 33.9 | |
| MNLI distributionBackbone=RoBERTa2024.12 | 31.7 | |
| Label-Free Explanation SelectionBackbone=RoBERTa, LLM=Llama3, Explanations=Model2024.12 | 31.4 | |
| MJD EstimatorBackbone=RoBERTa, LLM=GPT-4o, Explanations=None2024.12 | 30.9 | |
| Label-Guided Explanation SelectionBackbone=BERT, LLM=Llama3, Explanations=Model, Selection Strategy=VariErr Label-Guided2024.12 | 30.5 | |
| MJD EstimatorBackbone=RoBERTa, LLM=Llama3, Explanations=None2024.12 | 30.4 | |
| VariErr distributionBackbone=BERT2024.12 | 30.2 | |
| MJD EstimatorBackbone=BERT, LLM=Llama3, Explanations=Human2024.12 | 29.6 | |
| Label-Free Explanation SelectionBackbone=BERT, LLM=Llama3, Explanations=Model2024.12 | 29.2 | |
| MNLI-FT-LMBackbone=RoBERTa2024.12 | 29.2 | |
| Label-Free Explanation SelectionBackbone=BERT, LLM=GPT-4o, Explanations=Model2024.12 | 28.5 | |
| Label-Guided Explanation SelectionBackbone=BERT, LLM=Llama3, Explanations=Model, Selection Strategy=MNLI Label-Guided2024.12 | 28.4 | |
| ChaosNLI HJDBackbone=BERT2024.12 | 26.8 | |
| MJD EstimatorBackbone=BERT, LLM=GPT-4o, Explanations=None2024.12 | 25.8 | |
| MJD EstimatorBackbone=BERT, LLM=Llama3, Explanations=None2024.12 | 24.6 | |
| MNLI distributionBackbone=BERT2024.12 | 22.9 | |
| MNLI-FT-LMBackbone=BERT2024.12 | 22 | |
| Out-of-the-box LMBackbone=BERT2024.12 | 17 | |
| Out-of-the-box LMBackbone=RoBERTa2024.12 | 16.7 |