Text Watermarking in the Era of Large Language Models

We are excited to announce that our paper "A Survey of Text Watermarking in the Era of Large Language Models" has been accepted by ACM Computing Surveys. We would like to thank all the reviewers and editors for their valuable feedback and support.

• Added new works related to text watermarking from January to June 2024.
• Included more detailed mind maps (Figures 4, 6, 7, and 8) in sections on Watermarking for Existing Text and Watermarking for LLMs.
• Significant adjustments to the evaluation section:
  • Added a taxonomy (Figure 9) organizing all evaluation metrics.
  • Included Tables 2 and 3 for detailed analysis of metrics and attack algorithms.
  • Added Section 5.7 outlining key benchmarks and tools for text watermarking.
• Reorganized challenges and future directions section to focus more on technical challenges.

The latest version has been uploaded to arXiv.

Our survey is now available on arXiv.