Efficient, Direct, and Restricted Black-Box Graph Evasion Attacks to Any-Layer Graph Neural Networks via Influence Function

An official PyTorch implementation of "Efficient, Direct, and Restricted Black-Box Graph Evasion Attacks to Any-Layer Graph Neural Networks via Influence Function" (WSDM 2024). [paper]

Note

The code has not been organized yet. We will organize it as soon as possible. Thanks.

Running the code

For example, to check the performance of our one-time white box attack against GCN in cora dataset, run the following code:

cd OTA

python main.py --attack_object white --model GCN --dataset cora --attack_algorithm one-time