Mudjacking: Patching Backdoor Vulnerabilities in Foundation Models

arXiv:2402.14977v1 Announce Type: cross
Abstract: Foundation model has become the backbone of the AI ecosystem. In particular, a foundation model can be used as a general-purpose feature extractor to build various downstream classifiers. However, foundation models are vulnerable to backdoor attacks and a backdoored foundation model is a single-point-of-failure of the AI ecosystem, e.g., multiple downstream classifiers inherit the backdoor vulnerabilities simultaneously. In this work, we propose Mudjacking, the first method to patch foundation models to remove backdoors. Specifically, given …

abstract ai ecosystem arxiv attacks backdoor become build classifiers cs.cr cs.cv cs.lg ecosystem failure feature foundation foundation model general type vulnerabilities vulnerable