arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2510.01637cs.LG
收藏

基于组合水印的大语言模型水印输出生成后编辑检测

Detecting Post-generation Edits to Watermarked LLM Outputs via Combinatorial Watermarking

Department of Industrial and Systems Engineering, University of Minnesota(工业与系统工程系,明尼苏达大学);Google(谷歌);Department of Electrical and Computer Engineering, Princeton University(电气工程与计算机科学系,普林斯顿大学);Stony Brook University(石溪大学)

Liyan Xie, Muhammad Siddeek, Mohamed Seif, Andrea J. Goldsmith, Mengdi Wang

AI总结:

本研究针对带水印大语言模型输出的生成后编辑检测定位需求,提出基于组合模式的水印框架及配套全局、局部统计量,在多编辑场景下验证了其优异的编辑定位性能。

摘要:

水印技术已成为专有语言模型的核心技术,可用于区分AI生成文本与人类撰写文本。但在诸多现实场景中,大语言模型生成的内容可能会经历生成后编辑,比如人工修订甚至欺骗性攻击,因此检测并定位此类修改至关重要。本研究提出一项新任务:检测针对带水印大语言模型输出的局部生成后编辑。为此,我们提出一种基于组合模式的水印框架,该框架将词表划分为互不相交的子集,并在生成过程中通过在这些子集上强制执行确定性组合模式来嵌入水印。我们为组合水印搭配了可用于检测水印的全局统计量,此外还设计了轻量级的局部统计量,用于标记并定位潜在的编辑操作。我们提出了两项针对该任务的评估指标:I类错误率和检测准确率,并在多种编辑场景下基于开源大语言模型对方法进行了评估,结果表明该方法在编辑定位方面具备优异的实证性能。

英文摘要:

Watermarking has become a key technique for proprietary language models, enabling the distinction between AI-generated and human-written text. However, in many real-world scenarios, LLM-generated content may undergo post-generation edits, such as human revisions or even spoofing attacks, making it critical to detect and localize such modifications. In this work, we introduce a new task: detecting post-generation edits locally made to watermarked LLM outputs. To this end, we propose a combinatorial pattern-based watermarking framework, which partitions the vocabulary into disjoint subsets and embeds the watermark by enforcing a deterministic combinatorial pattern over these subsets during generation. We accompany the combinatorial watermark with a global statistic that can be used to detect the watermark. Furthermore, we design lightweight local statistics to flag and localize potential edits. We introduce two task-specific evaluation metrics, Type-I error rate and detection accuracy, and evaluate our method on open-source LLMs across a variety of editing scenarios, demonstrating strong empirical performance in edit localization.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑