Fine-tuning Pre-trained Language Models for Few-shot Intent Detection: Supervised Pre-training and Isotropization

21466

Knowledge Triples

Source Entity Relationship Target Entity
isBasedOn
isBasedOn
improves
improves