Mixture of basis for interpretable continual learning with distribution shifts

Open in new window