CEFW: A Comprehensive Evaluation Framework for Watermark in Large Language Models
This work addresses the problem of evaluating watermarking techniques for synthetic text detection, providing a standardized framework for researchers and practitioners, though it is incremental in nature.
The authors tackled the lack of unified evaluation for text watermarking in large language models by proposing CEFW, a comprehensive framework that assesses methods across five dimensions, and introduced Balanced Watermark, which outperforms existing methods in overall performance.
Text watermarking provides an effective solution for identifying synthetic text generated by large language models. However, existing techniques often focus on satisfying specific criteria while ignoring other key aspects, lacking a unified evaluation. To fill this gap, we propose the Comprehensive Evaluation Framework for Watermark (CEFW), a unified framework that comprehensively evaluates watermarking methods across five key dimensions: ease of detection, fidelity of text quality, minimal embedding cost, robustness to adversarial attacks, and imperceptibility to prevent imitation or forgery. By assessing watermarks according to all these key criteria, CEFW offers a thorough evaluation of their practicality and effectiveness. Moreover, we introduce a simple and effective watermarking method called Balanced Watermark (BW), which guarantees robustness and imperceptibility through balancing the way watermark information is added. Extensive experiments show that BW outperforms existing methods in overall performance across all evaluation dimensions. We release our code to the community for future research. https://github.com/DrankXs/BalancedWatermark.