Jason's work focuses on neuroevolution, deep learning, evolutionary algorithms, and computer vision. His current and past research interests include using evolutionary heuristics to design better deep neural architectures and bilevel neuroevolution.