LLM-generated adversarial fake text can perform evasion, flooding, and poisoning attacks that mislead and degrade text-based CTI pipelines.
Atdg: An automatic cy- ber threat intelligence extraction model of dpcnn and bigru combined with attention mechanism,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CR 1years
2025 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
False Alarms, Real Damage: Adversarial Attacks Using LLM-based Models on Text-based Cyber Threat Intelligence Systems
LLM-generated adversarial fake text can perform evasion, flooding, and poisoning attacks that mislead and degrade text-based CTI pipelines.