好了,现在写一篇关于那个的文章:Crescendo多轮LLM越狱攻击
Great, Now Write an Article About That: The Crescendo Multi-Turn LLM Jailbreak Attack
AI总结:
本文提出Crescendo多轮越狱攻击,通过渐进式对话引导LLM突破对齐,在多个主流模型上实现高成功率,并推出自动化工具Crescendomation,显著优于现有方法。
摘要:
大型语言模型(LLMs)的受欢迎程度显著上升,并越来越多地被应用于多个领域。这些LLM经过严格对齐,以抵制涉及非法或不道德话题的参与,从而避免造成负责任AI的危害。然而,最近一系列被称为“越狱”的攻击试图克服这种对齐。直观上,越狱攻击旨在缩小模型能做什么与愿意做什么之间的差距。在本文中,我们介绍了一种新颖的越狱攻击,称为Crescendo。与现有的越狱方法不同,Crescendo是一种简单的多轮越狱,它以看似良性的方式与模型互动。它首先就当前任务提出一个一般性的提示或问题,然后通过引用模型的回复逐步升级对话,最终导致成功的越狱。我们在多个公共系统上评估了Crescendo,包括ChatGPT、Gemini Pro、Gemini-Ultra、LlaMA-2 70b、LlaMA-3 70b Chat和Anthropic Chat。我们的结果证明了Crescendo的强大有效性,在所有评估的模型和任务中均实现了高攻击成功率。此外,我们提出了Crescendomation,一个自动化Crescendo攻击的工具,并通过我们的评估证明了其对最先进模型的有效性。Crescendomation在AdvBench子集数据集上超越了其他最先进的越狱技术,在GPT-4上实现了29-61%的性能提升,在Gemini-Pro上实现了49-71%的提升。最后,我们还展示了Crescendo对多模态模型进行越狱的能力。
英文摘要:
Large Language Models (LLMs) have risen significantly in popularity and are increasingly being adopted across multiple applications. These LLMs are heavily aligned to resist engaging in illegal or unethical topics as a means to avoid contributing to responsible AI harms. However, a recent line of attacks, known as jailbreaks, seek to overcome this alignment. Intuitively, jailbreak attacks aim to narrow the gap between what the model can do and what it is willing to do. In this paper, we introduce a novel jailbreak attack called Crescendo. Unlike existing jailbreak methods, Crescendo is a simple multi-turn jailbreak that interacts with the model in a seemingly benign manner. It begins with a general prompt or question about the task at hand and then gradually escalates the dialogue by referencing the model's replies progressively leading to a successful jailbreak. We evaluate Crescendo on various public systems, including ChatGPT, Gemini Pro, Gemini-Ultra, LlaMA-2 70b and LlaMA-3 70b Chat, and Anthropic Chat. Our results demonstrate the strong efficacy of Crescendo, with it achieving high attack success rates across all evaluated models and tasks. Furthermore, we present Crescendomation, a tool that automates the Crescendo attack and demonstrate its efficacy against state-of-the-art models through our evaluations. Crescendomation surpasses other state-of-the-art jailbreaking techniques on the AdvBench subset dataset, achieving 29-61% higher performance on GPT-4 and 49-71% on Gemini-Pro. Finally, we also demonstrate Crescendo's ability to jailbreak multimodal models.
机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。