Model merging is a promising lightweight model empowerment technique that
does not rely on expensive computing devices (e.g., GPUs) or require the
collection of specific training data. Instead, it involves editing different
upstream model parameters to absorb their downstream task capabilities.
However, uncertified model merging can infringe upon the Intellectual Property
(IP) rights of the original upstream models. In this paper, we conduct the
first study on the robustness of IP protection methods in model merging
scenarios. We investigate two state-of-the-art IP protection techniques:
Quantization Watermarking and Instructional Fingerprint, along with various
advanced model merging technologies, such as Task Arithmetic, TIES-MERGING, and
so on. Experimental results indicate that current Large Language Model (LLM)
watermarking techniques cannot survive in the merged models, whereas model
fingerprinting techniques can. Our research aims to highlight that model
merging should be an indispensable consideration in the robustness assessment
of model IP protection techniques, thereby promoting the healthy development of
the open-source LLM community.

模型合并是一种有前景的轻量级模型增强技术，它不依赖昂贵的计算设备（如 GPU），也不需要收集特定的训练数据。该技术通过编辑不同的上游模型参数，吸收它们的下游任务能力。本文研究了模型合并场景下知识产权（IP）保护方法的鲁棒性，并调查了量化水印和指令指纹等两种最先进的 IP 保护技术，以及各种先进的模型合并技术。实验结果表明，当前的大型语言模型（LLM）水印技术无法在合并模型中存活，而模型指纹技术可以。我们的研究旨在强调在模型 IP 保护技术的鲁棒性评估中应考虑模型合并，从而促进开源 LLM 社区的健康发展。