search

cmhcbb / Seq2Sick

Adversarial examples for Seq2Seq model in NLP
39 stars 5 forks source link

Proximal operator on group lasso #1

Closed gpengzhi closed 4 years ago

gpengzhi commented 4 years ago

Here

Why do you put the gradient updating part after the group lasso part instead of putting it before the group lasso part? I feel that you can not get a sparse solution at the end of the iterations. Please correct me if I am wrong.

Thanks a lot!