Lujia Shen (Zhejiang University), Yuwen Pu (Zhejiang University), Shouling Ji (Zhejiang University), Changjiang Li (Penn State), Xuhong Zhang (Zhejiang University), Chunpeng Ge (Shandong University), Ting Wang (Penn State)

Transformer-based models, such as BERT and GPT, have been widely adopted in natural language processing (NLP) due to their exceptional performance. However, recent studies show their vulnerability to textual adversarial attacks where the model's output can be misled by intentionally manipulating the text inputs. Despite various methods that have been proposed to enhance the model's robustness and mitigate this vulnerability, many require heavy consumption resources (e.g., adversarial training) or only provide limited protection (e.g., defensive dropout).

In this paper, we propose a novel method called dynamic attention, tailored for the transformer architecture, to enhance the inherent robustness of the model itself against various adversarial attacks. Our method requires no downstream task knowledge and does not incur additional costs. The proposed dynamic attention consists of two modules: *(i) attention rectification*, which masks or weakens the attention value of the chosen tokens, and *(ii) dynamic modeling*, which dynamically builds the set of candidate tokens. Extensive experiments demonstrate that dynamic attention significantly mitigates the impact of adversarial attacks, improving up to 33% compared to previous methods against widely used adversarial attacks. The model-level design of dynamic attention enables it to be easily combined with other defense methods (e.g., adversarial training) to further enhance the model's robustness. Furthermore, we demonstrate that dynamic attention preserves the state-of-the-art robustness space of the original model compared to other dynamic modeling methods.

View More Papers

BGP-iSec: Improved Security of Internet Routing Against Post-ROV Attacks

Cameron Morris (University of Connecticut), Amir Herzberg (University of Connecticut), Bing Wang (University of Connecticut), Samuel Secondo (University of Connecticut)

Read More

A Duty to Forget, a Right to be Assured?...

Hongsheng Hu (CSIRO's Data61), Shuo Wang (CSIRO's Data61), Jiamin Chang (University of New South Wales), Haonan Zhong (University of New South Wales), Ruoxi Sun (CSIRO's Data61), Shuang Hao (University of Texas at Dallas), Haojin Zhu (Shanghai Jiao Tong University), Minhui Xue (CSIRO's Data61)

Read More

WIP: Adversarial Object-Evasion Attack Detection in Autonomous Driving Contexts:...

Rao Li (The Pennsylvania State University), Shih-Chieh Dai (Pennsylvania State University), Aiping Xiong (Penn State University)

Read More

Understanding Route Origin Validation (ROV) Deployment in the Real...

Lancheng Qin (Tsinghua University, BNRist), Li Chen (Zhongguancun Laboratory), Dan Li (Tsinghua University, Zhongguancun Laboratory), Honglin Ye (Tsinghua University), Yutian Wang (Tsinghua University)

Read More