揭示大语言模型在越狱攻击中的说服性指纹
Uncovering the Persuasive Fingerprint of LLMs in Jailbreaking Attacks
AI总结：
本研究利用社会科学说服理论构建对抗性提示，发现大语言模型对说服性结构更顺从，并揭示其越狱响应中的独特说服性指纹，实证表明此类提示能显著绕过安全措施。
摘要：
尽管近期取得了进展，大语言模型仍然容易受到越狱攻击，这些攻击绕过对齐安全措施并引发有害输出。虽然先前的研究提出了各种在人类可读性和可迁移性方面有所不同的攻击策略，但很少关注可能影响模型对此类攻击敏感性的语言和心理机制。在本文中，我们研究了一条跨学科的研究路线，该路线利用社会科学中基础性的说服理论来构建能够规避大语言模型对齐约束的对抗性提示。基于公认的说服策略，我们假设经过大规模人类生成文本训练的大语言模型可能对具有说服性结构的提示做出更顺从的响应。此外，我们调查大语言模型本身是否在其越狱响应中表现出独特的说服性指纹。跨多个对齐大语言模型的实证评估表明，具有说服性意识的提示显著绕过安全措施，展示了其诱导越狱行为的潜力。这项工作强调了跨学科见解在应对大语言模型安全不断演变的挑战中的重要性。代码和数据已公开。
英文摘要：
Despite recent advances, Large Language Models remain vulnerable to jailbreak attacks that bypass alignment safeguards and elicit harmful outputs. While prior research has proposed various attack strategies differing in human readability and transferability, little attention has been paid to the linguistic and psychological mechanisms that may influence a model's susceptibility to such attacks. In this paper, we examine an interdisciplinary line of research that leverages foundational theories of persuasion from the social sciences to craft adversarial prompts capable of circumventing alignment constraints in LLMs. Drawing on well-established persuasive strategies, we hypothesize that LLMs, having been trained on large-scale human-generated text, may respond more compliantly to prompts with persuasive structures. Furthermore, we investigate whether LLMs themselves exhibit distinct persuasive fingerprints that emerge in their jailbreak responses. Empirical evaluations across multiple aligned LLMs reveal that persuasion-aware prompts significantly bypass safeguards, demonstrating their potential to induce jailbreak behaviors. This work underscores the importance of cross-disciplinary insight in addressing the evolving challenges of LLM safety. The code and data are available.
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。
