NLP

Pre-trained Language Models Return Distinguishable Probability Distributions to Unfaithfully Hallucinated Texts