GraphNAD uses degree-weighted graph attention transfer plus layer-relation congruence to distill backdoored GNNs on 3% clean data and lower attack success rate below 5%.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Fine-tuning is Not Fine: Mitigating Backdoor Attacks in GNNs with Limited Clean Data
GraphNAD uses degree-weighted graph attention transfer plus layer-relation congruence to distill backdoored GNNs on 3% clean data and lower attack success rate below 5%.