Reducing Oversmoothing through Informed Weight Initialization in Graph Neural Networks

Open in new window