RTE is a method where given a premise and hypotheses, the model has to predict if the premise entails the hypotheses.

A natural language inference task; typically solved by fine-tuning models like BERT.

References