Distilling the Thought, Watermarking the Answer: A Principle Semantic Guided Watermark for Large Reasoning Models

TL;DR

This paper presents ReasonMark, a novel semantic-guided watermarking framework for reasoning large language models that maintains logical coherence, enhances robustness, and incurs minimal latency overhead.

Contribution

The paper introduces a new watermarking approach that decouples reasoning and answering phases, using semantic importance to guide robust, low-latency watermark embedding in reasoning LLMs.

Findings

Reduces text perplexity by 0.35

Increases translation BLEU score by 0.164

Raises mathematical accuracy by 0.67 points

Abstract

Reasoning Large Language Models (RLLMs) excelling in complex tasks present unique challenges for digital watermarking, as existing methods often disrupt logical coherence or incur high computational costs. Token-based watermarking techniques can corrupt the reasoning flow by applying pseudo-random biases, while semantic-aware approaches improve quality but introduce significant latency or require auxiliary models. This paper introduces ReasonMark, a novel watermarking framework specifically designed for reasoning-intensive LLMs. Our approach decouples generation into an undisturbed Thinking Phase and a watermarked Answering Phase. We propose a Criticality Score to identify semantically pivotal tokens from the reasoning trace, which are distilled into a Principal Semantic Vector (PSV). The PSV then guides a semantically-adaptive mechanism that modulates watermark strength based on token-PSV alignment, ensuring robustness without compromising logical integrity. Extensive experiments show ReasonMark surpasses state-of-the-art methods by reducing text Perplexity by 0.35, increasing translation BLEU score by 0.164, and raising mathematical accuracy by 0.67 points. These advancements are achieved alongside a 0.34% higher watermark detection AUC and stronger robustness to attacks, all with a negligible increase in latency. This work enables the traceable and trustworthy deployment of reasoning LLMs in real-world applications.