@inproceedings{yanaka-etal-2025-intersectional,
    title = "Intersectional Bias in {J}apanese Large Language Models from a Contextualized Perspective",
    author = "Yanaka, Hitomi  and
      He, Xinqi  and
      Jie, Lu  and
      Han, Namgi  and
      Oh, Sunjin  and
      Kumon, Ryoma  and
      Matsuoka, Yuma  and
      Watabe, Kazuhiko  and
      Itatsu, Yuko",
    editor = "Fale{\'n}ska, Agnieszka  and
      Basta, Christine  and
      Costa-juss{\`a}, Marta  and
      Sta{\'n}czak, Karolina  and
      Nozza, Debora",
    booktitle = "Proceedings of the 6th Workshop on Gender Bias in Natural Language Processing (GeBNLP)",
    month = aug,
    year = "2025",
    address = "Vienna, Austria",
    publisher = "Association for Computational Linguistics",
    url = "https://preview.aclanthology.org/landing_page/2025.gebnlp-1.2/",
    doi = "10.18653/v1/2025.gebnlp-1.2",
    pages = "18--32",
    ISBN = "979-8-89176-277-0",
    abstract = "An growing number of studies have examined the social bias of rapidly developed large language models (LLMs). Although most of these studies have focused on bias occurring in a single social attribute, research in social science has shown that social bias often occurs in the form of intersectionality{---}the constitutive and contextualized perspective on bias aroused by social attributes. In this study, we construct the Japanese benchmark inter-JBBQ, designed to evaluate the intersectional bias in LLMs on the question-answering setting. Using inter-JBBQ to analyze GPT-4o and Swallow, we find that biased output varies according to its contexts even with the equal combination of social attributes."
}