Scalable and Order-robust Continual Learning with Additive Parameter Decomposition.

Jaehong Yoon, Saehoon Kim, Eunho Yang, Sung Ju Hwang

Browse the full ICLR paper archive.