DiGGR introduces a self-supervised graph representation learning framework that disentangles latent factors to guide mask modeling and improve representation quality on graph tasks.
Dataset IMDB-B IMDB-M PROTEINS COLLAB MUTAG REDDIT-B NCI1 Statistics Avg
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2024 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Disentangled Generative Graph Representation Learning
DiGGR introduces a self-supervised graph representation learning framework that disentangles latent factors to guide mask modeling and improve representation quality on graph tasks.