Shortcuts Arising from Contrast: Effective and Covert Clean-Label Attacks in Prompt-Based Learning

arXiv:2404.00461v1 Announce Type: new
Abstract: Prompt-based learning paradigm has demonstrated remarkable efficacy in enhancing the adaptability of pretrained language models (PLMs), particularly in few-shot scenarios. However, this learning paradigm has been shown to be vulnerable to backdoor attacks. The current clean-label attack, employing a specific prompt as a trigger, can achieve success without the need for external triggers and ensure correct labeling of poisoned samples, which is more stealthy compared to the poisoned-label attack, but on the other hand, it …

abstract adaptability arxiv attacks backdoor contrast cs.ai cs.cl cs.cr cs.lg current few-shot however language language models paradigm prompt prompt-based learning type vulnerable