Semantic Anomaly Detection on Warren Semantic
1AccuracyMasked language model
Evaluation results
| Method | Links | |
|---|---|---|
| Masked language modelBackbone=BERT, Anomaly Model=MLM, Layer Selection=Last layer2021.05 | 1 | |
| Masked language modelBackbone=RoBERTa, Anomaly Model=MLM, Layer Selection=Last layer2021.05 | 1 | |
| Gaussian anomaly modelBackbone=RoBERTa, Anomaly Model=GM, Layer Selection=Best layer2021.05 | 0.944 | |
| Gaussian anomaly modelBackbone=XLNet, Anomaly Model=GM, Layer Selection=Best layer2021.05 | 0.667 | |
| Gaussian anomaly modelBackbone=BERT, Anomaly Model=GM, Layer Selection=Best layer2021.05 | 0.556 | |
| Masked language modelBackbone=XLNet, Anomaly Model=MLM, Layer Selection=Last layer2021.05 | 0.556 |