RanDumb: A Simple Approach that Questions the Efficacy of Continual Representation Learning

Open in new window