Roland Zimmermann

Hi. Sadly, I don't fully understand your first question. Regarding your second question: please refer to the literature for this - a common choice is to choose the step size...

1. Are models expected to return logits or probabilities? (e.g. #618) 2. Which attacks currently exist and what modes/criteria do they support? (i.e. #653)

6. How can one use the implemented attacks with sigmoid networks, i.e. there is only a single output (see #664)?

Can you please elobrate a bit on your problem/question?

You are right that we should put this in the documentation. @jangop If you have the time and already an idea how to implement your suggested changes please feel free...

It looks like you normalize all samples as part of the `DataLoader` > transforms.Normalize([0.485, 0.456, 0.406], [0.229, 0.224, 0.225]) This means, that a the model does not operate in the...

I would just suggest updating the documentation to include this information.

Right now there is no implementation for this in foolbox. However, this should be pretty easy to implement. You could start by creating a class that inherits from the framework-specific...