QueerGen: How LLMs Reflect Societal Norms on Gender and Sexuality in Sentence Completion Tasks
Fuente:
arXiv
Enregistré dans:
| Auteurs principaux: | , , |
|---|---|
| Format: | Preprint |
| Publié: |
2026
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866910003624935424 |
|---|---|
| author | Sosto, Mae Pandiani, Delfina Sol Martinez Hollink, Laura |
| author_facet | Sosto, Mae Pandiani, Delfina Sol Martinez Hollink, Laura |
| contents | This paper examines how Large Language Models (LLMs) reproduce societal norms, particularly heterocisnormativity, and how these norms translate into measurable biases in their text generations. We investigate whether explicit information about a subject's gender or sexuality influences LLM responses across three subject categories: queer-marked, non-queer-marked, and the normalized "unmarked" category. Representational imbalances are operationalized as measurable differences in English sentence completions across four dimensions: sentiment, regard, toxicity, and prediction diversity. Our findings show that Masked Language Models (MLMs) produce the least favorable sentiment, higher toxicity, and more negative regard for queer-marked subjects. Autoregressive Language Models (ARLMs) partially mitigate these patterns, while closed-access ARLMs tend to produce more harmful outputs for unmarked subjects. Results suggest that LLMs reproduce normative social assumptions, though the form and degree of bias depend strongly on specific model characteristics, which may redistribute, but not eliminate, representational harms. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2601_20731 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | QueerGen: How LLMs Reflect Societal Norms on Gender and Sexuality in Sentence Completion Tasks Sosto, Mae Pandiani, Delfina Sol Martinez Hollink, Laura Computation and Language Artificial Intelligence Computers and Society This paper examines how Large Language Models (LLMs) reproduce societal norms, particularly heterocisnormativity, and how these norms translate into measurable biases in their text generations. We investigate whether explicit information about a subject's gender or sexuality influences LLM responses across three subject categories: queer-marked, non-queer-marked, and the normalized "unmarked" category. Representational imbalances are operationalized as measurable differences in English sentence completions across four dimensions: sentiment, regard, toxicity, and prediction diversity. Our findings show that Masked Language Models (MLMs) produce the least favorable sentiment, higher toxicity, and more negative regard for queer-marked subjects. Autoregressive Language Models (ARLMs) partially mitigate these patterns, while closed-access ARLMs tend to produce more harmful outputs for unmarked subjects. Results suggest that LLMs reproduce normative social assumptions, though the form and degree of bias depend strongly on specific model characteristics, which may redistribute, but not eliminate, representational harms. |
| title | QueerGen: How LLMs Reflect Societal Norms on Gender and Sexuality in Sentence Completion Tasks |
| topic | Computation and Language Artificial Intelligence Computers and Society |
| url | https://arxiv.org/abs/2601.20731 |