arXiv cs.CL论文
面向机器生成文本鲁棒水印检测的稳定性感知特征设计
针对大语言模型生成文本的水印检测在多次改写和短文本下性能下降的问题，本文提出模式稳定性评分（PSS）框架，结合局部统计特征、游程模式高阶统计、自相关信号及跨改写深度的稳定性评分，在PG-19、CNN/DailyMail和WikiText数据集上，使用Llama-3-8B、Qwen2-7B等模型和多种改写器进行测试，相比基线方法AUC提升10-15个百分点，且单一通用分类器可跨模型、改写器和领域泛化，AUC保持87.8%以上。
这篇是正式发表的长论文，站内提供中文解读，全文请到原文阅读 PDF。
Abstract:The widespread adoption of large language models (LLMs) has intensified the demand for principled methods to distinguish human from machine-generated text. Watermarking provides a promising avenue, yet existing detectors exhibit sharp performance deterioration under multiple paraphrasing and when applied to shorter texts. We introduce Pattern Stability Score (PSS), a novel detection framework that leverages local statistical features and stability dynamics across paraphrased variants. Specifically, the proposed method combines global and local z-score features with higher-order statistics of run-length patterns, enriched by autocorrelation signals and stability scores computed over paraphrase depth. Numerical evaluations are performed on three benchmark datasets (PG-19, CNN/DailyMail, and WikiText) using multiple LLMs (Llama-3-8B, Qwen2-7B) and paraphrasers (Mistral-7B, Qwen2-7B, Gemma-7B), systematically stress-testing robustness under up to eight rounds of paraphrasing. Compared to prior z-score thresholding baselines and some state-of-the-art deep learning methods, our approach improves detection AUC (area under the receiver operating characteristic curve) by over 10-15 percentage points across different token lengths. Additionally, extensive cross-domain experiments demonstrate that a single universal classifier generalizes across different LLMs, paraphrasers, and text domains without retraining, maintaining above 87.8% AUC even when all components differ from training.