.png){kind=icon}
返回
SecNLP: An NLP classification model watermarking framework based on multi-task learning
.png){kind=icon}
PRE
AI
DOI：10.1016/j.csl.2023.101606.png)
.png){kind=icon}
原文求助.png){kind=icon}
分享.png){kind=icon}
收藏
摘要
En 中文
The popularity of ChatGPT demonstrates the immense commercial value of natural language processing (NLP) technology. However, NLP models like ChatGPT are vulnerable to piracy and redistribution, which can harm the economic interests of model owners. Existing NLP model watermarking schemes struggle to balance robustness and covertness. Typically, robust watermarks require embedding more information, which compromises their covertness; conversely, covert watermarks are challenging to embed more information, which affects their robustness. This paper is proposed to use multi-task learning (MTL) to address the conflict between robustness and covertness. Specifically, a covert trigger set is established to implement remote verification of the watermark model, and a covert auxiliary network is designed to enhance the watermark model's robustness. The proposed watermarking framework is evaluated on two benchmark datasets and three mainstream NLP models. Compared with existing schemes, the framework not only has excellent covertness and robustness but also has a lower false positive rate and can effectively resist fraudulent ownership claims by adversaries.
Keyword:
Natural language processing
NLP model security
Black-box watermarking
White-box watermarking
期刊
C
IF:
3.4
论文数:
1.5K
被引数:
2.6K
机构
引用论文
Darwin’s Tree of Life is Numbered. Resolving the Origins of Species by Mass达尔文的生命之树已被编号。通过质量解析物种的起源。
PREAI
分享收藏A Backdoor Attack Against LSTM-Based Text Classification Systems一种针对基于LSTM的文本分类系统的后门攻击.png)
IEEE ACCESS
OAAI
IEEE ACCESSIF3.6
OAAI分享收藏
The Diagnostic Utility of a Commercial Limb-Girdle Muscular Dystrophy Gene Test Panel商业性肢体带肌营养不良基因检测板的诊断效用
PREAI
分享收藏Coming undone: hemiparasite presence and effects in a prairie grassland diminish over timeOecologia
OecologiaIF0
PREAI
分享收藏
查看更多内容