Tag
The paper introduces CTQW-GNN, a graph neural network leveraging continuous-time quantum walks to address over-smoothing and improve performance on heterophilic graphs, demonstrating state-of-the-art results on benchmark datasets.