Shreyashee Sinha

Evaluating Paraphrastic Robustness in Textual Entailment Models

We present PaRTE, a collection of 1,126 pairs of Recognizing Textual Entailment (RTE) examples to evaluate whether models are robust to paraphrasing. We posit that if RTE models understand language, their predictions should be consistent across inputs that share the same meaning. We use the evaluati…