Mole-BERT: Rethinking Pre-training Graph Neural Networks for Molecules
Mole-BERT: Rethinking Pre-training Graph Neural Networks for Molecules
Jun Xia,Chengshuai Zhao,5 Authors,Stan Z. Li
2023 · DBLP: conf/iclr/XiaZHG0LLL23
International Conference on Learning Representations · 178 Citations
TLDR
A novel node-level pre-training task, dubbed Masked Atoms Modeling (MAM), to mask some discrete codes randomly and then pre-train GNNs to predict them and mitigates another issue of AttrMask, namely the negative transfer.