高效定位大文档中被水印标记的小片段，兼顾速度与准确率。
WaterSeeker: Pioneering Efficient Detection of Watermarked Segments in Large Documents
- 先快速提取可疑区域，再局部精确检测。
- 在真实场景下实现高精度定位，计算开销更低。
- 适合需要可解释性的AI内容审核系统使用。
大型语言模型（LLMs）的水印算法在检测生成文本方面已达到高准确率。然而，现有方法主要关注区分完全水印文本与非水印文本，忽视了实际场景中仅部分段落由LLM生成的情况。此时，需在时间复杂度与检测性能间取得平衡，面临显著挑战。本文提出WaterSeeker，一种高效检测并定位大文档中水印片段的新方法。首先采用高效的异常提取技术初步定位可疑水印区域；随后进行局部遍历，并对重点区域执行全文检测以实现更精确验证。理论分析与实验结果表明，WaterSeeker在检测准确率与计算效率之间实现了更优平衡。其定位能力为构建可解释的AI检测系统奠定了基础。代码已公开于 https://github.com/THU-BPM/WaterSeeker。
原文摘要 · Abstract (English)
Watermarking algorithms for large language models (LLMs) have attained high accuracy in detecting LLM-generated text. However, existing methods primarily focus on distinguishing fully watermarked text from non-watermarked text, overlooking real-world scenarios where LLMs generate only small sections within large documents. In this scenario, balancing time complexity and detection performance poses significant challenges. This paper presents WaterSeeker, a novel approach to efficiently detect and locate watermarked segments amid extensive natural text. It first applies an efficient anomaly extraction method to preliminarily locate suspicious watermarked regions. Following this, it conducts a local traversal and performs full-text detection for more precise verification. Theoretical analysis and experimental results demonstrate that WaterSeeker achieves a superior balance between detection accuracy and computational efficiency. Moreover, its localization capability lays the foundation for building interpretable AI detection systems. Our code is available at https://github.com/THU-BPM/WaterSeeker.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。