无刻板印象的个性化：语音大语言模型是否知道何时该“充耳不闻”？
Personalization Without Stereotyping: Do Speech LLMs Know When to Turn a Deaf Ear?
AI总结：
本研究针对语音大语言模型的刻板印象问题，构建BASIS-Bench基准评估8款模型，发现其虽能识别人口统计学线索却难实现恰当个性化，明确标签会放大刻板印象，强调需训练模型合理利用相关线索。
摘要：
近期的语音大语言模型（Speech LLMs）可利用丰富的副语言线索，实现更复杂、个性化的交互。然而，鲜有研究探讨这种敏感性如何导致刻板印象，现有评估大多仅衡量人口统计学差异，未区分有益的个性化与刻板印象行为。我们推出BASIS-Bench基准，旨在通过三个维度拆解这两方面：（1）人口统计学鲁棒性，衡量基于内容的任务在不同说话人及人口统计学分类能力上的差异；（2）刻板印象偏差，衡量感知到的人口统计学线索是否会触发无提示的刻板印象行为；（3）个性化，衡量模型是否能在明确要求时根据人口统计学特征调整响应。我们评估了8款近期的语音大语言模型，发现尽管多数模型常能识别人口统计学线索，但大多未能将其转化为恰当的个性化；明确的人口统计学标签可提升个性化，但也会放大文本主干中已存在的刻板印象关联。这些发现凸显，训练语音大语言模型不仅要感知人口统计学线索，还要学会何时让这些线索影响自身响应。
英文摘要：
Recent speech LLMs can leverage rich paralinguistic cues, enabling more complex and personalized interactions. However, little work investigates how this sensitivity can lead to stereotyping, and existing evaluations largely measure demographic disparities without distinguishing beneficial personalization from stereotypical behavior. We introduce BASIS-Bench, a benchmark designed to disentangle these two aspects through three axes: (1) demographic robustness, measuring disparities in content-based tasks across speakers and demographic classification capabilities; (2) stereotypical bias, measuring whether perceived demographic cues trigger unprompted stereotypical behavior; and (3) personalization, measuring whether models can tailor responses to demographic characteristics when explicitly requested. We evaluate eight recent speech LLMs and find that, despite often recognizing demographic cues, most fail to translate them into appropriate personalization. Explicit demographic labels improve personalization but also amplify stereotypical associations already present in the textual backbones. These findings highlight the need to train speech LLMs not only to perceive demographic cues, but to learn when they should shape their responses.
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。
