Less Is More -- On the Importance of Sparsification for Transformers and Graph Neural Networks for TSP

Open in new window