Return to Issue Details Accelerating Graph Transformer Models through GPU-Aware Kernel Fusion Download Download PDF