ClockMark: Adaptive and Host-Dependent Watermarking for Large Language Models

A new arXiv paper (2610.27079v1) proposes ClockMark, an LLM watermarking framework that embeds the full payload in each watermarked token and, the authors prove, preserves the original model's sequence distribution. Across experiments against 12 baselines, ClockMark improved robustness to text editing by up to 16.08 percentage points in bit accuracy over the strongest tested baseline. The method rests on two principles the authors argue for: adaptive embedding that retains prior sampling decisions, and host-dependent embedding that uses the current LLM distribution to constrain token selection and state updates.

arXiv:2610.27079v1 Announce Type: new Abstract: Large language model LLM watermarking operates on the next-token distribution that evolves during generation. In streaming generation, each sampled token becomes part of the text and affects subsequent distributions. Embedding therefore proceeds with a changing host, unlike classical watermarking, which adapts embedding to an existing host signal. To accommodate this difference, we formulate a novel embedding framework for LLM watermarking based on two principles. We first argue that embedding should be adaptive, retaining the state of previous sampling decisions to better guide subsequent choices. We further argue that embedding should be host-dependent, using the current LLM distribution to constrain both token selection and state updates. Guided by these principles and the two requirements of distribution preservation and robustness to text editing, we propose ClockMark. It combines adaptive and host-dependent embedding with message spreading, embedding the full payload in each watermarked token. Theoretically, we prove that ClockMark preserves the original LLM sequence distribution, thereby maintaining generation quality. Empirically, extensive experiments against 12 baselines demonstrate improved robustness to text editing, with gains of up to 16.08 percentage points in bit accuracy over the strongest tested baseline.