paper to implement
reference[30]
This is the paper that I am implementing in PyTorch and am unable to understand how to compute this term for days now. I understand that Mu and sigma are both 300D learned vectors and s being the semantic representation of the label ( which is again a 300-D word2vec vector ). Can anyone explain as to how the three terms are relating each other in this equation as this notation is weird to me?
Thanks in advance for your help…