Introduces graph-label conditioned (GLC) and embedding-label conditioned (ELC) reconstruction attacks on GNNs that achieve high-quality graph recovery in black-box settings on NCI1, PROTEINS and AIDS datasets using four distributional metrics.
Model inversion attacks through target-specific conditional diffusion models,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.AI 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Rethinking Generative Reconstruction Attacks against Graph Neural Network Models
Introduces graph-label conditioned (GLC) and embedding-label conditioned (ELC) reconstruction attacks on GNNs that achieve high-quality graph recovery in black-box settings on NCI1, PROTEINS and AIDS datasets using four distributional metrics.