Generate a tailored SOP for Dr. Arthur Jacot. Improve your application with a focused, well-structured draft.
Arthur Jacot is an Assistant Professor at the Courant Institute of Mathematical Sciences at New York University. Previously, he was a PhD student at École Polytechnique Fédérale de Lausanne under the supervision of Clément Hongler. His research focuses on developing new mathematical concepts and tools to describe the training dynamics of Deep Neural Networks (DNNs) while creating theories in Deep Learning. Currently, he is excited about a project that applies a computational version of Occam's razor to find the fastest algorithms for fitting training data with DNNs. His interests include feature learning, particularly the emergence of low-dimensional representations, weight decay, and identifying regimes in DNN training, aiming to develop a complete phase diagram capturing diverse training dynamics. He actively invites students from NYU to contact him via email if they are interested in potential projects that align with their mathematical and empirical inquiries.
Open Program in Biomedical Sciences (Vilcek Institute) covers departments like Biochemistry, Pathology, Neuroscience, Microbiology, etc.