Excited to share that our work regarding the privacy leakage of adversarial training models has been accepted to The Art of Robustness workshop at CVPR as an oral presentation! See [html][code][arxiv] for more details.