Trainability, Expressivity and Interpretability in Gated Neural ODEs

Open in new window