ATAG-LLM uses LLM-generated text and a similarity proxy to inject nodes that mislead GNN classifiers without touching embeddings.
Adversarial attack and defense on graph data: A survey,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.AI 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Navigating the Black Box: Leveraging LLMs for Effective Text-Level Graph Injection Attacks
ATAG-LLM uses LLM-generated text and a similarity proxy to inject nodes that mislead GNN classifiers without touching embeddings.