5 comments

[ 3.2 ms ] story [ 33.0 ms ] thread
Another thin layer over a word2vec model.
I think it might not even do that. From the title, looks like it just cares about words being used in the same sentence, not in similar contexts.