The Language Interpretability Tool (LIT) is a visual, interactive model-understanding tool for NLP models.
