Zahra Kadkhodaie is a PhD student of Data Science at NYU (advised by Prof. Eero Simoncelli) and a guest researcher in the Center for Computational Neuroscience. Her research focuses on understanding and improving deep neural networks by analyzing and imposing mathematical symmetries on the architecture. More recently, she has worked on extracting and utilizing the prior embedded in a trained deep neural network denoiser for solving other computer vision problems without further training. Zahra completed her B.Sc. in Solid State Physics at K.N. Toosi University in Tehran, Iran and her M.Sc in Data Science and Psychology at NYU.