提问方式的重要性：大语言模型（LLMs）中与性别相关的语言偏差
It's How You Ask: Gender-Associated Linguistic Bias in LLMs
浏览论文内容
中文总结 AI 辅助
该研究发现LLMs会因提示中女性常用的语言特征生成更短欠正式的回应，其影响远大于明确性别线索，且事后缓解困难，呼吁关注语言变异以缓解LLM介导职场交流的差异化影响。
中文摘要 AI 辅助
专业交流越来越多地由大语言模型（LLMs）介导，但这些模型是否平等服务于所有用户？我们表明，当提示包含女性更常用的语言特征（模糊限制语、附加疑问句、集体指称）时，在三种文档类型和四个模型中，它们会系统地引发更短、欠复杂且欠正式的回应。在控制提示复杂度和特征传递后，这些效应仍然存在。诸如署名等明确的性别线索被编码在与语言方言相同的表征空间中，暗示存在共同的潜在机制，但语言语域的影响要大得多，会产生显著且一致的效应，而姓名则无此效应。我们的结果进一步表明，事后缓解具有挑战性：由于这些模式嵌入文化且不受意识控制，用户难以通过策略性自我呈现轻易避免它们，且机制分析显示，语言特征被编码在Transformer的早期层中并与其他特征纠缠在一起。我们的工作呼吁从上游考虑语言变异的影响，以缓解LLM介导的职场交流的差异化影响。
英文摘要
Professional communication is increasingly mediated by LLMs - but do these models serve all users equally? We show that when prompts contain linguistic features more commonly used by women (hedges, tag questions, collective reference), they systematically elicit shorter, less sophisticated, and less formal responses across three document types and four models. These effects persist after controlling for prompt complexity and feature carry-over. Explicit gender cues like sign-off names are encoded in the same representational space as linguistic dialect - suggesting shared underlying mechanisms - yet linguistic register is far more influential, producing large, consistent effects where names produce none. Our results further reveal that post-hoc mitigation is challenging: because these patterns are culturally embedded and outside conscious control, users cannot easily avoid them through strategic self-presentation, and mechanistic analysis reveals that linguistic features are encoded in early transformer layers and entangled with other features. Our work calls for upstream consideration of the influences of linguistic variation to mitigate disparate impacts of LLM-mediated workplace communication.
