EMNLP Findings2023
Watermarking LLMs with Weight Quantization
Linyang Li, Botian Jiang, Pengyu Wang, Ke Ren, Hang Yan, Xipeng Qiu
Abstract
Abuse of large language models reveals high risks as large language models are being deployed at an astonishing speed. It is important to protect the model weights to avoid malicious usage that violates licenses of opensource large language models. This paper proposes a novel watermarking strategy that plants watermarks in the quantization process of large language models without pre-defined triggers during inference. The watermark works when the model is used in the fp32 mode and remains hidden when the model is quantized to int8, in this way, the users can only inference the model without further supervised fine-tuning of the model. We successfully plant the watermark into open-source large language model weights including GPT-Neo and LLaMA. We hope our proposed method can provide a potential direction for protecting model weights in the era of large language model applications. 1
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.
6 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
- BitMark: Watermarking Bitwise Autoregressive Image Generative Models cites A Watermark for Large Language Models
- DetectGPT: Zero-Shot Machine-Generated Text Detection using Probability Curvature cites Language Models are Few-Shot Learners
- DetectGPT: Zero-Shot Machine-Generated Text Detection using Probability Curvature cites A Watermark for Large Language Models
- ErrorTrace: A Black-Box Traceability Mechanism Based on Model Family Error Space cites EmMark: Robust Watermarks for IP Protection of Embedded Quantized Large Language Models
- ErrorTrace: A Black-Box Traceability Mechanism Based on Model Family Error Space cites Language Models are Few-Shot Learners
- ErrorTrace: A Black-Box Traceability Mechanism Based on Model Family Error Space cites A Watermark for Large Language Models
- Protecting Copyright of Medical Pre-trained Language Models: Training-Free Backdoor Model Watermarking cites EmMark: Robust Watermarks for IP Protection of Embedded Quantized Large Language Models
- Protecting Copyright of Medical Pre-trained Language Models: Training-Free Backdoor Model Watermarking cites Language Models are Few-Shot Learners
- WaterDrum: Watermark-based Data-centric Unlearning Metric cites A Watermark for Large Language Models
Cited by top-tier papers6
- EmMark: Robust Watermarks for IP Protection of Embedded Quantized Large Language ModelsRuisi Zhang, Farinaz KoushanfarDAC 2024 · 11 citations
- WaterDrum: Watermark-based Data-centric Unlearning MetricXinyang Lu, Xinyuan Niu, Gregory Kang Ruey Lau, Nhung Bui et al.ICLR 2026 · 8 citations
- BitMark: Watermarking Bitwise Autoregressive Image Generative ModelsLouis Kerner, Michel Meintz, Bihe Zhao, Franziska Boenisch et al.NeurIPS 2025 · 6 citations
- ErrorTrace: A Black-Box Traceability Mechanism Based on Model Family Error SpaceChuanchao Zang, Xiangtao Meng, Wenyu Chen, Tianshuo Cong et al.NeurIPS 2025 · 4 citations
- Protecting Copyright of Medical Pre-trained Language Models: Training-Free Backdoor Model WatermarkingCong Kong, Rui Xu, Jiawei Chen, Zhaoxia YinACM MM 2025 · 1 citation
Builds on10
- Language Models are Few-Shot LearnersTom B. Brown, Benjamin Mann, Nick Ryder, Melanie Subbiah et al.NeurIPS 2020 · 64,255 citations
- DetectGPT: Zero-Shot Machine-Generated Text Detection using Probability CurvatureEric Mitchell, Yoonho Lee, Alexander Khazatsky, Christopher D. Manning et al.ICML 2023 · 988 citations
- A Watermark for Large Language ModelsJohn Kirchenbauer, Jonas Geiping, Yuxin Wen, Jonathan Katz et al.ICML 2023 · 854 citations
- Q-BERT: Hessian Based Ultra Low Precision Quantization of BERTSheng Shen, Zhen Dong, Jiayu Ye, Linjian Ma et al.AAAI 2020 · 656 citations
- Weight Poisoning Attacks on Pretrained ModelsKeita Kurita, Paul Michel, Graham NeubigACL 2020 · 312 citations
Related papers
- WaterMax: breaking the LLM watermark detectability-robustness-quality trade-offEva Giboulot, Teddy FuronNeurIPS 2024 · 76 citations
- Are You Copying My Model? Protecting the Copyright of Large Language Models for EaaS via Backdoor WatermarkWenjun Peng, Jingwei Yi, Fangzhao Wu, Shangxi Wu et al.ACL 2023 · 39 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
- Provable Robust Watermarking for AI-Generated TextXuandong Zhao, Prabhanjan Vijendra Ananth, Lei Li, Yu-Xiang WangICLR 2024 · 312 citations
- An Efficient White-box LLM Watermarking for IP Protection on Online Market PlatformsShuguang Yuan, Xingyu Su, Peizhuo Lv, Weiji Xue et al.KDD 2025
