A Principled Approach to Natural Language Watermarking
Zhe Ji, Qiansiqi Hu, Yicheng Zheng, Liyao Xiang, Xinbing Wang
2024年份
4被引次数
1顶会引用
摘要
Recently, there has been a surge in machine-generated natural language content being misused by unauthorized parties. Watermarking is a well-recognized technique to address the issue by tracing the provenance of the text. However, we found that most existing watermarking systems for texts are subject to ad hoc design and thus suffer from fundamental vulnerabilities.
问问这篇 Paper
问问你的智能体。
Lune 读过与它相关的顶会 Paper,每个回答都会注明依据哪几篇。
引用它的顶会 Paper1
问问它们各自怎么用它相关 Paper
- Tracing Text Provenance via Context-Aware Lexical SubstitutionXi Yang, Jie Zhang, Kejiang Chen, Weiming Zhang 等AAAI 2022 · 被引用 89 次
- No Free Lunch in LLM Watermarking: Trade-offs in Watermarking Design ChoicesQi Pang, Shengyuan Hu, Wenting Zheng, Virginia SmithNeurIPS 2024 · 被引用 56 次
- Provably Robust Multi-bit Watermarking for AI-generated TextWenjie Qu, Wengrui Zheng, Tianyang Tao, Dong Yin 等USENIX Security 2025
- Segmenting Watermarked Texts From Language ModelsXingchi Li, Guanxun Li, Xianyang ZhangNeurIPS 2024 · 被引用 5 次
- Linear Ensembles Wash Away Watermarks: On the Fragility of Distributional Perturbations in LLMsZhihao Wu, Gracia Gong, Qinglin Zhu, Yudong Chen 等ICML 2026
