Grammatical Acceptability on CoLA (held-out)
46.8F1 ScoreExpNet
Evaluation results
| Method | Links | |
|---|---|---|
| ExpNetTraining Data=SST-2 + HateXplain, Evaluation Type=Cross-dataset2026.01 | 46.8 | |
| RandomBaselineTraining Data=SST-2 + HateXplain, Evaluation Type=Cross-dataset2026.01 | 38.7 | |
| GradCAMTraining Data=SST-2 + HateXplain, Evaluation Type=Cross-dataset2026.01 | 35.6 | |
| LRPTraining Data=SST-2 + HateXplain, Evaluation Type=Cross-dataset2026.01 | 35.5 | |
| CAMTraining Data=SST-2 + HateXplain, Evaluation Type=Cross-dataset2026.01 | 35.5 | |
| GAETraining Data=SST-2 + HateXplain, Evaluation Type=Cross-dataset2026.01 | 35.4 | |
| MGAETraining Data=SST-2 + HateXplain, Evaluation Type=Cross-dataset2026.01 | 35.4 | |
| RawAtTraining Data=SST-2 + HateXplain, Evaluation Type=Cross-dataset2026.01 | 35.3 | |
| RolloutTraining Data=SST-2 + HateXplain, Evaluation Type=Cross-dataset2026.01 | 34.7 | |
| AttCATTraining Data=SST-2 + HateXplain, Evaluation Type=Cross-dataset2026.01 | 34.5 | |
| Integrated GradientTraining Data=SST-2 + HateXplain, Evaluation Type=Cross-dataset2026.01 | 34.2 | |
| FullLRPTraining Data=SST-2 + HateXplain, Evaluation Type=Cross-dataset2026.01 | 33.6 | |
| SHAPTraining Data=SST-2 + HateXplain, Evaluation Type=Cross-dataset2026.01 | 33 | |
| LIMETraining Data=SST-2 + HateXplain, Evaluation Type=Cross-dataset2026.01 | 32.3 |