Please tell me about the attack algorithms compared in the article.

[pdsu-zjl]

I understand from your excellent article that you have reproduced a lot of noise based adversarial attack methods. I would like to ask you if there are any adversarial attacks that can be achieved without labelling by reproducing those attack methods. That is, the dataset has no labels, only GT graphs, and I would like to ask you if you have any attack algorithms for this kind of dataset. Or I would like to ask you if there is a way to apply those attack methods with this kind of dataset.

Thank you.