Universal Vulnerabilities in Large Language Models: Backdoor Attacks for In-context Learning

arXiv:2401.05949v4 Announce Type: replace
Abstract: In-context learning, a paradigm bridging the gap between pre-training and fine-tuning, has demonstrated high efficacy in several NLP tasks, especially in few-shot settings. Despite being widely applied, in-context learning is vulnerable to malicious attacks. In this work, we raise security concerns regarding this paradigm. Our studies demonstrate that an attacker can manipulate the behavior of large language models by poisoning the demonstration context, without the need for fine-tuning the model. Specifically, we design a new …

abstract arxiv attacks backdoor concerns context cs.ai cs.cl few-shot fine-tuning gap in-context learning language language models large language large language models nlp paradigm pre-training raise security security concerns tasks training type vulnerabilities vulnerable work