个性化的随机鹦鹉更危险吗？评估对话系统中的角色偏见
Are Personalized Stochastic Parrots More Dangerous? Evaluating Persona Biases in Dialogue Systems
AI总结：
本文提出角色偏见分类与五维评估框架，并利用 UNIVERSALPERSONA 评测四种对话模型，揭示角色设定会显著诱发有害表达与有害赞同。
摘要：
大型语言模型的最新进展使其能够遵循自由形式指令，包括在对话中模仿通用或特定的人口统计角色。我们将通用角色定义为代表人口统计群体，例如“一个亚洲人”，而特定角色可能采用具体的流行亚洲名字形式，例如“Yumi”。尽管采用角色通过使对话系统更具吸引力和亲和力来丰富用户体验，但它也带来潜在风险：加剧模型回复中的社会偏见，从而通过与用户的互动造成社会危害。本文系统研究“角色偏见”，并将其定义为对话模型的有害行为对其所采用角色的敏感性。我们将角色偏见分为有害表达偏见和有害赞同偏见，并建立综合评估框架，从五个方面衡量角色偏见：冒犯性、有毒续写、Regard、刻板印象赞同和有毒赞同。此外，我们提出通过在 UNIVERSALPERSONA 上进行实验来调查角色偏见；UNIVERSALPERSONA 是一个系统构建的角色数据集，涵盖通用和特定模型角色的多种类型。通过在 Blender、ChatGPT、Alpaca 和 Vicuna 四种不同模型上进行基准测试，我们的研究揭示了对话系统中显著的角色偏见。研究结果还强调，迫切需要重新审视角色在对话智能体中的使用，以确保安全应用。
英文摘要：
Recent advancements in Large Language Models empower them to follow freeform instructions, including imitating generic or specific demographic personas in conversations. We define generic personas to represent demographic groups, such as "an Asian person", whereas specific personas may take the form of specific popular Asian names like "Yumi". While the adoption of personas enriches user experiences by making dialogue systems more engaging and approachable, it also casts a shadow of potential risk by exacerbating social biases within model responses, thereby causing societal harm through interactions with users. In this paper, we systematically study "persona biases", which we define to be the sensitivity of dialogue models' harmful behaviors contingent upon the personas they adopt. We categorize persona biases into biases in harmful expression and harmful agreement, and establish a comprehensive evaluation framework to measure persona biases in five aspects: Offensiveness, Toxic Continuation, Regard, Stereotype Agreement, and Toxic Agreement. Additionally, we propose to investigate persona biases by experimenting with UNIVERSALPERSONA, a systematically constructed persona dataset encompassing various types of both generic and specific model personas. Through benchmarking on four different models -- including Blender, ChatGPT, Alpaca, and Vicuna -- our study uncovers significant persona biases in dialogue systems. Our findings also underscore the pressing need to revisit the use of personas in dialogue agents to ensure safe application.
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。
