Closed anianruoss closed 6 years ago
That's a difference in notation between Eq. (3) of Xiao et al. and Carlini & Wagner (see bottom of p. 9 in https://arxiv.org/pdf/1608.04644.pdf). Xiao et al. assume a negative kappa (well in practice they only consider kappa = 0) while Carlini & Wagner assume a positive one. As the documentation of the function adv_loss
primarily refers to the Carlini & Wagner paper I think it is sufficiently clear.
Thank you for clarifying!
In my opinion, line 115 in losses.py should be:
instead of:
according to eq. (3) in Xiao et al.
Am I missing something?