Prediction rules in deep learning are based on a forward, recursive computation through several layers. Implicit deep learning rules go much beyond, by relying on the solution of an implicit (or, “fixed-point”) equation that has to be numerically solved in order to make the prediction: for... Read more