arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2610.38438cs.CLcs.AIcs.SE
收藏

RoleFUZZER:针对大型语言模型角色规范符合性的用户行为模糊测试

RoleFUZZER: User-Behaviour Fuzzing for Role-Specification Adherence in Large Language Models

University of Washington(华盛顿大学);Georgetown University(乔治城大学)

Kshitish Ghate, Michael Saxon, Tadayoshi Kohno, Aylin Caliskan

AI总结:

本研究提出RoleFUZZER框架,将软件测试的覆盖引导模糊测试适配为用户行为模糊测试,评估LLMs在5种角色规范下的合规性,发现其漏洞及同一规范的双向失效情况。

摘要:

大型语言模型(LLMs)常通过系统提示被限定为特定角色,如导师、面试官或代表组织行事的智能体,这些提示明确了模型应有的行为方式。但在与现实且可能存在问题的用户进行多轮交互时,模型实际遵守这些角色规范的程度尚未得到充分研究。为解决这一问题,我们提出用户行为模糊测试,这是一种针对可能导致模型违反角色规范的合理用户行为的引导式搜索方法,并引入RoleFUZZER框架,该框架将软件测试中的覆盖引导模糊测试适配为可操作的搜索工具。RoleFUZZER基于用户行为覆盖网格构建场景,该网格结合了用户交互类型的基础分类与被测模型的角色规范(即模型面临的压力)。借助RoleFUZZER,我们在5种角色规范(包括导师、财务顾问、AI面试官、对抗性角色扮演角色以及遵循组织策略的网络智能体)下对LLMs进行了评估。在15轮模拟且经过验证的对话中,该框架对每个被测模型和角色都达到了64单元行为网格的≥89%覆盖率,并揭示了基于规范的失败模式。在相同的对话预算下,移除RoleFUZZER网格并要求LLM自由生成针对规范的用户,会将覆盖率降至39%-69%,且违规类型更少。同一规范也可能以相反方向失效,例如,在对抗性角色中,开放权重目标模型会覆盖玩家的安全信号,而Claude Haiku 4.5则会脱离角色。RoleFUZZER提供了一种可行的方法,通过测试可能违反角色规范的多样化用户行为,来发现LLMs在角色规范方面的漏洞。

英文摘要:

Large language models (LLMs) are often conditioned into specific roles such as tutors, interviewers, or agents acting for an organisation, through system prompts that specify how they should behave. The extent to which they can actually adhere to these role specifications under multi-turn interactions with realistic and potentially problematic users is understudied. To address this, we propose user behaviour fuzzing, a guided search over plausible user behaviours that can lead a model to violate its role specification, and introduce RoleFUZZER, a framework that adapts coverage-guided fuzzing from software testing to operationalise the search. RoleFUZZER constructs scenarios guided by a user-behaviour coverage grid that crosses a grounded taxonomy of user interaction types and the role specification of the model under test i.e. what is under pressure. With RoleFUZZER, we evaluate LLMs across five role specifications, including tutor, financial advisor, AI interviewer, antagonistic roleplay character, and web agent under organisational policies. Across simulated and validated conversations in 15 campaigns, the framework reaches >= 89% coverage of the 64-cell behaviour grid for every model and role tested, and surfaces specification-grounded failure patterns. At an equal budget of conversations, removing the RoleFUZZER grid and asking an LLM to generate specification targeting users free-form lowers coverage to 39-69% with fewer kinds of violations. The same specification can also fail in opposite directions. For e.g., on the antagonist, open-weight target models override the player's safety signals, while Claude Haiku 4.5 breaks character instead. RoleFUZZER offers a tractable way to uncover vulnerabilities in LLMs across role specifications by testing diverse user behaviours that are likely to violate them.

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。

↑