ACL Findings2024
Subtle Signatures, Strong Shields: Advancing Robust and Imperceptible Watermarking in Large Language Models
Yubing Ren, Ping Guo, Yanan Cao, Wei Ma
Abstract
The widespread adoption of Large Language Models (LLMs) has led to an increase in AIgenerated text on the Internet, presenting a crucial challenge to differentiate AI-created content from human-written text. This challenge is critical to prevent issues of authenticity, trust, and potential copyright violations. Current research focuses on watermarking LLMgenerated text, but traditional techniques struggle to balance robustness with text quality. We introduce a novel watermarking approach, Robust and Imperceptible Watermarking (RIW) for LLMs, which leverages token prior probabilities to improve detectability and maintain watermark imperceptibility. RIW methodically embeds watermarks by partitioning selected tokens into two distinct groups based on their prior probabilities and employing tailored strategies for each group. In the detection stage, RIW method employs the 'voted z-test' to provide a statistically robust framework to identify the presence of a watermark accurately. The effectiveness of RIW is evaluated across three key dimensions: success rate, text quality, and robustness against removal attacks. Our experimental results on various LLMs, including GPT2-XL, OPT-1.3B, and LLaMA2-7B, indicate that RIW surpasses existing models, and also exhibits increased robustness against various attacks and good imperceptibility, thus promoting the responsible use of LLMs. Our code is available at https://github.com/Lilice- r/RIW.
Ask Lune
Know this paper in minutes, not an afternoon.
Lune can read every page of it and the top-tier papers that cite it, and gives you its claims, its evidence and its limits, each one cited to the paper it comes from. No skimming, no AI slop.
2 top-tier papers have cited it since. Know what they found before your reviewers do.
Lune can draw on
the full paper · the top-tier papers that cite it · related work
Free to start. No credit card required.
Lineage
- A Semantic Invariant Robust Watermark for Large Language Models cites Paraphrasing evades detectors of AI-generated text, but retrieval is an effective defense
- Paraphrasing evades detectors of AI-generated text, but retrieval is an effective defense cites A Semantic Invariant Robust Watermark for Large Language Models
- A Semantic Invariant Robust Watermark for Large Language Models cites Provable Robust Watermarking for AI-Generated Text
- Paraphrasing evades detectors of AI-generated text, but retrieval is an effective defense cites Provable Robust Watermarking for AI-Generated Text
- Provable Robust Watermarking for AI-Generated Text cites Paraphrasing evades detectors of AI-generated text, but retrieval is an effective defense
- Paraphrasing evades detectors of AI-generated text, but retrieval is an effective defense cites Training language models to follow instructions with human feedback
- Provable Robust Watermarking for AI-Generated Text cites Training language models to follow instructions with human feedback
- Unbiased Watermark for Large Language Models cites Paraphrasing evades detectors of AI-generated text, but retrieval is an effective defense
- From Trade-off to Synergy: A Versatile Symbiotic Watermarking Framework for Large Language Models cites Paraphrasing evades detectors of AI-generated text, but retrieval is an effective defense
- From Trade-off to Synergy: A Versatile Symbiotic Watermarking Framework for Large Language Models cites Provable Robust Watermarking for AI-Generated Text
- From Trade-off to Synergy: A Versatile Symbiotic Watermarking Framework for Large Language Models cites Unbiased Watermark for Large Language Models
- Towards Reliable Marking and Verification of AI-Generated Text via Geometry-aware Sentence-level Watermarking cites A Semantic Invariant Robust Watermark for Large Language Models
- Towards Reliable Marking and Verification of AI-Generated Text via Geometry-aware Sentence-level Watermarking cites Provable Robust Watermarking for AI-Generated Text
- Towards Reliable Marking and Verification of AI-Generated Text via Geometry-aware Sentence-level Watermarking cites From Trade-off to Synergy: A Versatile Symbiotic Watermarking Framework for Large Language Models
Cited by top-tier papers2
- From Trade-off to Synergy: A Versatile Symbiotic Watermarking Framework for Large Language ModelsYidan Wang, Yubing Ren, Yanan Cao, Binxing FangACL 2025 · 4 citations
- Towards Reliable Marking and Verification of AI-Generated Text via Geometry-aware Sentence-level WatermarkingYubing Ren, Ping Guo, Yanan CaoICML 2026
Builds on9
- Training language models to follow instructions with human feedbackLong Ouyang, Jeffrey Wu, Xu Jiang, Diogo Almeida et al.NeurIPS 2022 · 24,707 citations
- Paraphrasing evades detectors of AI-generated text, but retrieval is an effective defenseKalpesh Krishna, Yixiao Song, Marzena Karpinska, John Wieting et al.NeurIPS 2023 · 657 citations
- Provable Robust Watermarking for AI-Generated TextXuandong Zhao, Prabhanjan Vijendra Ananth, Lei Li, Yu-Xiang WangICLR 2024 · 312 citations
- A Semantic Invariant Robust Watermark for Large Language ModelsAiwei Liu, Leyi Pan, Xuming Hu, Shiao Meng et al.ICLR 2024 · 108 citations
- Unbiased Watermark for Large Language ModelsZhengmian Hu, Lichang Chen, Xidong Wu, Yihan Wu et al.ICLR 2024 · 103 citations
Related papers
- Adaptive Text Watermark for Large Language ModelsYepeng Liu, Yuheng BuICML 2024 · 63 citations
- Token-Specific Watermarking with Enhanced Detectability and Semantic Coherence for Large Language ModelsMingjia Huo, Sai Ashish Somayajula, Youwei Liang, Ruisi Zhang et al.ICML 2024 · 37 citations
- Watermarking with Low-Entropy POS-Guided Token Partitioning and Z-Score-Driven Dynamic Bias for Large Language ModelsHe Li, Xiaojun Chen, Zhendong Zhao, Yunfei Yang et al.EMNLP Findings 2025
- PASA: A Principled Embedding-Space Watermarking Approach for LLM-Generated Text under Semantic-Invariant AttacksZhenxin Ai, Haiyun HeICML 2026 · 4 citations
- WaterMax: breaking the LLM watermark detectability-robustness-quality trade-offEva Giboulot, Teddy FuronNeurIPS 2024 · 76 citations
