On the N-gram Approximation of Pre-trained Language Models

Re-assign community