all AI news
DrAttack: Prompt Decomposition and Reconstruction Makes Powerful LLM Jailbreakers
Feb. 28, 2024, 5:49 a.m. | Xirui Li, Ruochen Wang, Minhao Cheng, Tianyi Zhou, Cho-Jui Hsieh
cs.CL updates on arXiv.org arxiv.org
Abstract: The safety alignment of Large Language Models (LLMs) is vulnerable to both manual and automated jailbreak attacks, which adversarially trigger LLMs to output harmful content. However, current methods for jailbreaking LLMs, which nest entire harmful prompts, are not effective at concealing malicious intent and can be easily identified and rejected by well-aligned LLMs. This paper discovers that decomposing a malicious prompt into separated sub-prompts can effectively obscure its underlying malicious intent by presenting it in …
abstract alignment arxiv attacks automated cs.ai cs.cl cs.cr current jailbreak jailbreaking language language models large language large language models llm llms prompt prompts safety type vulnerable
More from arxiv.org / cs.CL updates on arXiv.org
Jobs in AI, ML, Big Data
Senior Machine Learning Engineer
@ GPTZero | Toronto, Canada
ML/AI Engineer / NLP Expert - Custom LLM Development (x/f/m)
@ HelloBetter | Remote
Doctoral Researcher (m/f/div) in Automated Processing of Bioimages
@ Leibniz Institute for Natural Product Research and Infection Biology (Leibniz-HKI) | Jena
Coding Data Quality Auditor
@ Neuberger Berman | Work At Home-Georgia
Post Graduate (Year-Round) Intern - Market Research Analyst and Agreement Support
@ National Renewable Energy Laboratory | CO - Golden
Retail Analytics Engineering - Sr. Manager (Data)
@ Axalta | Woonsocket-1 CVS Drive