让扩散语言模型生成带水印的文本，靠调整解码顺序实现
dgMARK: Decoding-Guided Watermarking for Diffusion Language Models
- 通过控制解码顺序引导特定位置生成符合奇偶校验的词
- 水印检测依赖奇偶匹配统计，对增删改等编辑仍有效
- 无需修改模型即可兼容常见解码策略，适合内容安全场景
我们提出 dgMARK，一种针对离散扩散语言模型（dLLMs）的解码引导型水印方法。与自回归模型不同，dLLMs 可以任意顺序生成词元。尽管理想条件下条件预测器应与掩码顺序无关，但实际 dLLMs 对去掩码顺序敏感，这为水印提供了新通道。dgMARK 通过引导去掩码顺序，使高奖励候选词满足由二值哈希诱导的简单奇偶性约束，而无需显式重加权模型学习概率。该方法可即插即用，兼容常见解码策略（如置信度、熵、边际基排序），并可通过一步前瞻变体增强效果。水印通过提升的奇偶匹配统计量进行检测，滑动窗口检测器确保在插入、删除、替换和改写等后处理操作下仍具鲁棒性。
原文摘要 · Abstract (English)
We propose dgMARK, a decoding-guided watermarking method for discrete diffusion language models (dLLMs). Unlike autoregressive models, dLLMs can generate tokens in arbitrary order. While an ideal conditional predictor would be invariant to this order, practical dLLMs exhibit strong sensitivity to the unmasking order, creating a new channel for watermarking. dgMARK steers the unmasking order toward positions whose high-reward candidate tokens satisfy a simple parity constraint induced by a binary hash, without explicitly reweighting the model's learned probabilities. The method is plug-and-play with common decoding strategies (e.g., confidence, entropy, and margin-based ordering) and can be strengthened with a one-step lookahead variant. Watermarks are detected via elevated parity-matching statistics, and a sliding-window detector ensures robustness under post-editing operations including insertion, deletion, substitution, and paraphrasing. Project website: https://dgmark-watermarking.github.io
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。