Request edit access
NLP. Quiz 2
Simple Word Vector representations: word2vec, GloVe
What type of functions can approximate neural network with linear activation function?
Only linear functions
Only continuous functions
If a vector representation for a word “jungle” is [-0.123 0.432 1.453 -0.003]. What seems like a vector representation for its synonym “forest” in the same vector space? Check all right answers.
[0 0 0 0 0 0 0 1]
[-0.120 0.410 1.312 -0.012]
[0 0 0 1 0 1]
[-0.140 0.5 1.479 0.002]
[-1.453 0.002 0.132 -0.231]
What seems like a reasonable expectation for a word in a pair “Russia - … “ for a Glove model trained on a news corpus where pairs “USA - Washington” and “France - Paris” are clearly traceable?
What are the possible ways for sentence embedding construction?
Suppose we train Softmax classifier over word embeddings for classification problem with `m` classes. `V` is vocabulary size, `d` is dimension of word embedding, `m` is number of classes. How many elements we will have in `W` -- weight matrix for Softmax classifier?
Never submit passwords through Google Forms.
This content is neither created nor endorsed by Google.
Terms of Service