UNK usually comes from the final sampling step: the distribution of words contains a special token UNK to represent all words with insufficient statistics in th

UNK usually comes from the final sampling step: the distribution of words contains a special token UNK to represent all words with insufficient statistics in the training corpus.