all InfoSec news
Have You Merged My Model? On The Robustness of Large Language Model IP Protection Methods Against Model Merging
April 9, 2024, 4:11 a.m. | Tianshuo Cong, Delong Ran, Zesen Liu, Xinlei He, Jinyuan Liu, Yichen Gong, Qi Li, Anyu Wang, Xiaoyun Wang
cs.CR updates on arXiv.org arxiv.org
Abstract: 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 …
arxiv collection computing cs.ai cs.cl cs.cr data devices editing empowerment gpus ip protection language large large language model protection robustness training training data upstream
More from arxiv.org / cs.CR updates on arXiv.org
Jobs in InfoSec / Cybersecurity
Cybersecurity Consultant
@ Devoteam | Cité Mahrajène, Tunisia
GTI Manager of Cybersecurity Operations
@ Grant Thornton | Phoenix, AZ, United States
(Senior) Director of Information Governance, Risk, and Compliance
@ SIXT | Munich, Germany
Information System Security Engineer
@ Space Dynamics Laboratory | North Logan, UT
Intelligence Specialist (Threat/DCO) - Level 3
@ Constellation Technologies | Fort Meade, MD
Cybersecurity GRC Specialist (On-site)
@ EnerSys | Reading, PA, US, 19605