arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2605.26397cs.CLcs.AI
收藏

LLM生成的自闭症交流中的算法脆弱性与人格偏见

Algorithmic Fragility and Persona Bias in LLM-Generated Autistic Communication

University of California, San Diego(加州大学圣地亚哥分校);Georgia Institute of Technology(佐治亚理工学院);Cornell University(康奈尔大学);Cardiff University(卡迪夫大学)

Naba Rizvi, Mohammed Rizvi, Harper Strickland, Saleha Ahmedi, Nedjma Ousidhoum

AI总结:

通过双人格改写范式,发现LLM在生成自闭症人格文本时存在词汇与情感偏离、输出坍塌等系统性失败,且对齐策略而非参数规模主导这些失败,表明当前对齐训练导致深层表征鸿沟。

摘要:

安全对齐减少了明确有害的输出,但无意中编码了一种对边缘化交流的净化、神经典型化表征。我们使用双人格改写范式研究这种编码,提示十个大型语言模型(LLM)从自闭症或神经典型人格改写自然发生的自闭症话语。我们发现,尽管语义相似性相当,自闭症人格改写比神经典型改写在词汇形式和情感语域上偏离显著更多。此外,大多数模型将跨人格生成折叠成几乎相同的输出。为了揭示这种生成崩溃背后的机制,我们引入了一个多智能体定性分析框架。我们的结果揭示了系统性输出擦除、刻板幻觉和任务回避元评论是此任务的普遍失败模式,这些模式按对齐策略而非参数规模聚类。最后,我们与自闭症人类标注者的针对性比较表明,社区内部知识相对于LLM分类产生了系统性标签反转。我们的发现表明,当前的对齐训练导致仅通过定性分析可见的人格特定生成崩溃,证实了提示工程无法解决的深层表征鸿沟。

英文摘要:

Safety alignment reduces explicitly harmful outputs but inadvertently encodes a sanitized, neuronormative representation of marginalized communication. We investigate this encoding using a dual-persona rewrite paradigm, prompting ten large language models (LLMs) to rewrite naturally occurring autistic discourse from either an autistic or neurotypical persona. We uncover autistic-persona rewrites diverge significantly more in lexical form and affective register than neurotypical rewrites, despite equivalent semantic similarity. Furthermore, most models collapse cross-persona generations into near-identical outputs. To uncover the mechanisms behind this generative breakdown, we introduce a multi-agent qualitative analysis framework. Our results reveal systemic output erasure, stereotyped hallucination, and task-evasive meta-commentary are pervasive failure modes for this task that cluster by alignment strategy rather than parameter scale. Finally, our targeted comparison with autistic human annotators demonstrates that community-insider knowledge produces systematic label reversals relative to LLM classifications. Our findings indicate that current alignment training causes persona-specific generative breakdown visible only through qualitative analysis, confirming a deep representational gap that prompt engineering cannot resolve.

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。

↑