dgMARK: Decoding-Guided Watermarking for Diffusion Language Models
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
Code (0)
등록된 구현이 없습니다.
Similar Papers 제목 키워드 기반
DMark: Order-Agnostic Watermarking for Diffusion Large Language Models
Diffusion large language models (dLLMs) offer faster generation than autoregressive models while maintaining comparable quality, but existing watermarking methods fail on them due to their non-sequential decoding. Unlike…
Diffusion-Based Image Editing for Breaking Robust Watermarks
Robust invisible watermarking aims to embed hidden information into images such that the watermark can survive various image manipulations. However, the rise of powerful diffusion-based image generation and editing techn…
Image GenerationImage EditingWorldCup Sampling for Multi-bit LLM Watermarking
As large language models (LLMs) generate increasingly human-like text, watermarking has emerged as a promising solution for reliable attribution beyond mere detection. While multi-bit watermarking enables richer provenan…
SAC-Copula: Quality-Preserving Watermarking for Diffusion Language Models via Smooth Correlated Gumbel Fields
Watermarking diffusion language models (DLMs) requires mechanisms compatible with iterative parallel unmasking rather than autoregressive decoding. Existing sampling-based watermarking methods typically inject position-w…
On the Learnability of Watermarks for Language Models
Watermarking of language model outputs enables statistical detection of model-generated text, which can mitigate harms and misuses of language models. Existing watermarking strategies operate by altering the decoder of a…
DecoderLanguage ModelingLanguage Modelling