deep-rules
deep-rules copied to clipboard
Overall discussion for Tip 10
This is to discuss outstanding issues for Tip 10: Don’t share models trained on sensitive data. https://github.com/Benjamin-Lee/deep-rules/blob/master/content/12.privacy.md
I think we need to make it clearer that it's not only that you shouldn't share the model weights (white box attacks) but even allow people to submit their own data and share the prediction (black box attacks).