Alert button
Picture for Lujia Shen

Lujia Shen

Alert button

Improving the Robustness of Transformer-based Large Language Models with Dynamic Attention

Nov 30, 2023
Lujia Shen, Yuwen Pu, Shouling Ji, Changjiang Li, Xuhong Zhang, Chunpeng Ge, Ting Wang

Viaarxiv icon

TextDefense: Adversarial Text Detection based on Word Importance Entropy

Feb 12, 2023
Lujia Shen, Xuhong Zhang, Shouling Ji, Yuwen Pu, Chunpeng Ge, Xing Yang, Yanghe Feng

Figure 1 for TextDefense: Adversarial Text Detection based on Word Importance Entropy
Figure 2 for TextDefense: Adversarial Text Detection based on Word Importance Entropy
Figure 3 for TextDefense: Adversarial Text Detection based on Word Importance Entropy
Figure 4 for TextDefense: Adversarial Text Detection based on Word Importance Entropy
Viaarxiv icon

Backdoor Pre-trained Models Can Transfer to All

Oct 30, 2021
Lujia Shen, Shouling Ji, Xuhong Zhang, Jinfeng Li, Jing Chen, Jie Shi, Chengfang Fang, Jianwei Yin, Ting Wang

Figure 1 for Backdoor Pre-trained Models Can Transfer to All
Figure 2 for Backdoor Pre-trained Models Can Transfer to All
Figure 3 for Backdoor Pre-trained Models Can Transfer to All
Figure 4 for Backdoor Pre-trained Models Can Transfer to All
Viaarxiv icon