Depen Morwani

dmorwani@g.harvard.edu, depenmorwani@gmail.com

Hi everyone!! I am a 3rd year PhD student in the ML Foundations group at Harvard, advised by Boaz Barak and Sham Kakade. I am broadly interested in optimization and generalization related aspects of deep learning. My current research focuses on understanding the inductive bias of optimization algorithms in deep learning, and utilizing it for coming up with better (in terms of efficiency/speed/bias towards safe/fair/robust features) optimization algorithms. Previously, I completed my Masters from IIT Madras under the guidance of Harish Guruprasad, where we worked on understanding the implications of weight normalization. I have also spent an year as a Research Associate at Google Research India, working with Praneeth Netrapalli and Prateek Jain on exploring the simplicity bias induced by gradient descent.