Mazda Moayeri studies the interpretability and robustness of machine learning models, with an emphasis on understanding the theoretical foundations of deep networks.