arXiv：cs.CL· Jiska Beuk, Gerasimos Spanakis·· 7 小时前AI 评分41
WinoQueer-NL：评估荷兰语模型对 LGBTQ+ 群体的偏见
WinoQueer-NL: Assessing Bias in Dutch Language Models toward LGBTQ+ Identities
AI 导读
研究者基于英文 WinoQueer 基准构建了荷兰语数据集 WinoQueer-NL，含 42,906 句，并经 43 名荷兰酷儿参与者在线调查验证。评估显示模型平均偏见分看似中性（约 50%），但部分模型对跨性别身份有高达 97% 的时间偏向刻板印象句，而男同性恋相关句仅 6%，跨性别与非二元身份得分最高。
正文
Abstract:While English language models have been widely examined for anti-queer bias, Dutch models remain understudied. To address this gap, we developed a culturally and linguistically adapted Dutch dataset based on the English WinoQueer benchmark, containing pairs of stereotypical and counter-stereotypical sentences. To validate and expand it, we conducted an online survey with 43 Dutch queer participants, confirming 145 of 171 stereotypes as culturally relevant and identifying 22 new biases through free-text responses. The final released dataset, comprising 42,906 sentences, was evaluated using a range of Dutch-specific and multilingual models, including both masked language models (MLMs) and autoregressive language models (ARLMs), with bias measured via a score comparing log-likelihoods of stereotypical versus counter-stereotypical sentences. While the mean bias score across models appeared neutral (~50%), closer analysis revealed significant disparities: some models favored stereotypical sentences up to 97% of the time for transgender identities, but only 6% of the time for gay-related pairs, with transgender and non-binary identities consistently receiving the highest bias scores. Our findings highlight the importance of culturally grounded datasets for evaluating and mitigating biases that disproportionately impact marginalized groups in Dutch language models.
| Comments: | accepted at 7th Workshop on Gender Bias in Natural Language Processing (GeBNLP2026) @ AACL2026. Dataset available via this https URL |
| Subjects: | Computation and Language (cs.CL) |
| Cite as: | arXiv:2609.02651 [cs.CL] |
| (or arXiv:2609.02651v2 [cs.CL] for this version) | |
| https://doi.org/10.48550/arXiv.2609.02651 arXiv-issued DOI via DataCite |
Submission history
From: Gerasimos Spanakis [view email]
[v1]
Wed, 2 Sep 2026 14:25:50 UTC (79 KB)
[v2]
Tue, 6 Oct 2026 14:58:17 UTC (82 KB)
来源：arXiv：cs.CL · arxiv.org