Organizingrecurrentnetworkdynamicsby task-computationtoenablecontinuallearning

Here, we develop a novel learning rule designed to minimize interferencebetween sequentially learned tasksinrecurrent networks.