My research interests are broadly in representation learning and how models learn useful internal representations from data. I am interested in using these representations to build deep learning models that are accurate, interpretable, and reliable.