.png){kind=icon}
返回
Transferable adversarial examples can efficiently fool topic models
.png){kind=icon}
PRE
AI
DOI：10.1016/j.cose.2022.102749.png)
.png){kind=icon}
原文求助.png){kind=icon}
分享.png){kind=icon}
收藏
摘要
En 中文
Intelligent text systems have been shown vulnerable to adversarial attacks. A popular way to enhance the robustness is to leverage transferable adversarial examples. As one of the most important statistical models, topic model is widely used in various applications. However, whether adversarial examples exhibit good transferability across topic models has never been explored. In this paper, we propose an effective adversarial example generator TopicAttack to disturb the inference of a target victim topic model and an ensemble algorithm TopicAttack+ to further increase transferability. TopicAttack exploits the words of high importance and constructs adversarial examples via synonym transformation. TopicAttack+ searches the optimal model ensemble given a set of prescribed substitutes, then generate adversarial examples of stronger transferability. Numerically, we demonstrate the effectiveness of the proposed method on the benchmark datasets AP, NIPS and 20News. In particular, given eight benchmark substitute models, TopicAttack disturbs them all effectively by achieving high Kullback-Leibler (KL) divergence; and TopicAttack+ significantly improves the average transferability performance on NIPS from 1.308 to 1.695. (C) 2022 Elsevier Ltd. All rights reserved.
Keyword:
Topic model
Adversarial attack
Transferability
KL divergence
Ensemble attack
期刊
C
IF:
5.4
论文数:
4.6K
被引数:
1.4W
机构
引用论文
.png){kind=icon}
Enhanced Reactivity of Dinuclear Copper(I) Acetylides in Dipolar Cycloadditions偶极环加成反应中双核乙炔铜 (I) 的反应性增强
PREAI
分享收藏
分享收藏
查看更多内容