Table 5

Results on PCBA. The evaluation metric is the Average Precision (AP↑). Bold values indicate the best performance.

ModelYearSourceTest-AP (%) ↑
DeeperGCN-VN-FLAG [20]2020arXiv preprint, ICML28.42±0.43
DGN [1]2021ICML28.85±0.30
GINE-VN [2, 6]2020arXiv preprint, ICML29.17±0.15
PHC-GNN [19]2021ICANN29.47±0.26
GIN-VN [6, 38]2018ICLR, ICML29.02±0.17
Graphormer-FLAG [40]2021NeurIPS31.40±0.34
TGT-Ag+TGT-At-DP [13]2024ICML31.67±0.31
HieGT  33.25 ± 0.27

or Create an Account

Close Modal
Close Modal