R-CoT: A Reasoning-Layer Watermark via Redundant Chain-of-Thought in Large Language Models
For LLM owners, this provides a watermarking method that is robust against output-level perturbations and post-training modifications, addressing a key vulnerability of existing superficial watermarking techniques.
R-CoT embeds watermarks into the reasoning path of LLMs via a redundant chain-of-thought framework, achieving high robustness with TPR above 95% under fine-tuning and other post-training operations.
Large language models (LLMs) are widely deployed in multiple scenarios due to reasoning capabilities. In order to prevent the models from being misused, watermarking is generally employed to ensure ownership. However, most existing watermarking methods rely on superficial modifications to the model's output distribution, rendering the watermark vulnerable to perturbation and removal. To overcome this challenge, this paper introduces a reasoning-layer framework termed Redundant Chain-of-Thought (R-CoT), which embeds watermarks into the reasoning path. A dual-trajectory optimization mechanism based on GRPO enables the native and the watermark reasoning path to coexist within a shared parameter space, internalizing the watermark as a distinct reasoning policy. Therefore, the watermark is embedded into the model's stable reasoning path, avoiding the watermark failure caused by output-level perturbations. Experimental results show that, compared with existing methods, R-CoT achieves high watermark effectiveness and strong robustness. Under fine-tuning and other post-training operations, the true positive rate (TPR) consistently remains above 95%, exhibiting only marginal degradation.