Demystifying Oversmoothing in Attention-Based Graph Neural Networks

Open in new window