Distilling Task-specific Logical Rules from Large Pre-trained Models

Open in new window