Syntactic Ghost: An Imperceptible General-purpose Backdoor Attacks on Pre-trained Language Models

arXiv:2402.18945v1 Announce Type: cross
Abstract: Pre-trained language models (PLMs) have been found susceptible to backdoor attacks, which can transfer vulnerabilities to various downstream tasks. However, existing PLM backdoors are conducted with explicit triggers under the manually aligned, thus failing to satisfy expectation goals simultaneously in terms of effectiveness, stealthiness, and universality. In this paper, we propose a novel approach to achieve invisible and general backdoor implantation, called \textbf{Syntactic Ghost} (synGhost for short). Specifically, the method hostilely manipulates poisoned samples with …

abstract arxiv attacks backdoor cs.ai cs.cl cs.cr found general ghost language language models tasks terms transfer type vulnerabilities