Eigenvector Dynamics of Kernel Methods @ ICLR 2023
Machine Learning
We've been working on large-scale kernel based machine learning methods that are competitive with deep neural networks. My work on developing an accessible framework for observing the training dynamics of these models and analyzing them was accepted at ICLR 2023 with an invited presentation.