[TMLR'24] Official code for "ProFeAT: Projected Feature Adversarial Training for Self-Supervised Learning of Robust Representations". - View it on GitHub