Which of the following word embedding architectures uses the current word to
predict the context?
Which of the following uses the context to predict the missing word in the middle?
Which of the following considers the importance of a word across the whole
document when predicting?
Which of the following ignores grammar and word order?