1 paper · 2 filters
Jędrzej Kozal, Michał Woźniak
Purpose: We propose a novel method for continual learning based on the increasing depth of neural networks. This work explores whether extending neural network depth may be benefic…