Next Level Message-Passing with Hierarchical Support Graphs

Benchmark Model Rank Results
graph-classification-on-peptides-funcGatedGCN-HSG#18AP: 0.6866±0.0038
graph-property-prediction-on-ogbg-molpcbaGatedGCN-HSG#4Test AP: 0.3129±0.0020
graph-regression-on-peptides-structGatedGCN-HSG#2MAE: 0.2421±0.0007
node-classification-on-coco-spGatedGCN-HSG#4macro F1: 0.3535±0.0032
node-classification-on-pascalvoc-sp-1GatedGCN-HSG#2macro F1: 0.4604±0.0059