Backdoors Against Natural Language Processing: A Review

IEEE Security & Privacy(2022)

引用 18|浏览49
暂无评分
摘要
Data poisoning attacks, specifically backdoor attacks, present a severe security threat in artificial intelligence. We provide a comprehensive survey into state-of-the-art backdoor attacks and defenses in the field of natural language processing.
更多
查看译文
关键词
Transformers, Natural language processing, History, Pipelines, Tokenization, Recurrent neural networks, Internet
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要