Entailment as Few-Shot Learner

Sinong Wang, Han Fang, Madian Khabsa +2

Large pre-trained language models (LMs) have demonstrated remarkable ability as few-shot learners. However, their success hinges largely on scaling model parameters to a degree tha…