学会伪装：通过多智能体攻击者-伪装者博弈规避 LLM 防御中的拒绝回复
Learn to Disguise: Avoid Refusal Responses in LLM's Defense via a Multi-agent Attacker-Disguiser Game
AI总结：
针对大模型拒绝式防御易被攻击者识别的问题，本文提出多智能体攻击者-伪装者博弈框架，通过博弈策略优化和课程学习隐藏防御意图，并可适配任意黑盒大模型。
摘要：
随着大模型在自然语言处理任务上的性能增强，大模型潜在的道德与伦理问题也随之出现。存在恶意攻击者通过提示工程等技术诱导大模型越狱，并生成包含非法、侵犯隐私信息的内容。因此，大模型使用安全对齐等技术反击恶意攻击者的攻击。然而，大模型通过拒绝回复形成的强防御机制很容易被攻击者识别，并被用于增强攻击者的能力。本文提出一种多智能体攻击者-伪装者博弈方法，以实现弱防御机制，使大模型既能安全地回复攻击者，又能隐藏防御意图。首先，我们构建一个多智能体框架来模拟攻击与防御场景，由不同角色分别负责攻击、伪装、安全评估和伪装评估任务。随后，我们设计攻击与伪装博弈算法，以优化攻击者和伪装者的博弈策略，并使用课程学习流程增强智能体的能力。实验验证，与其他方法相比，本文方法在增强模型伪装防御意图的能力方面更加有效。此外，我们的方法能够适配任意黑盒大模型以辅助模型防御，并且不会受到模型版本迭代的影响。
英文摘要：
With the enhanced performance of large models on natural language processing tasks, potential moral and ethical issues of large models arise. There exist malicious attackers who induce large models to jailbreak and generate information containing illegal, privacy-invasive information through techniques such as prompt engineering. As a result, large models counter malicious attackers' attacks using techniques such as safety alignment. However, the strong defense mechanism of the large model through rejection replies is easily identified by attackers and used to strengthen attackers' capabilities. In this paper, we propose a multi-agent attacker-disguiser game approach to achieve a weak defense mechanism that allows the large model to both safely reply to the attacker and hide the defense intent. First, we construct a multi-agent framework to simulate attack and defense scenarios, playing different roles to be responsible for attack, disguise, safety evaluation, and disguise evaluation tasks. After that, we design attack and disguise game algorithms to optimize the game strategies of the attacker and the disguiser and use the curriculum learning process to strengthen the capabilities of the agents. The experiments verify that the method in this paper is more effective in strengthening the model's ability to disguise the defense intent compared with other methods. Moreover, our approach can adapt any black-box large model to assist the model in defense and does not suffer from model version iterations.
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。
