QueerGen: Evaluating representational harm toward queer identities in language models

Sosto, Mae

QueerGen is a framework for evaluating how large language models (LLMs) represent queer and non-normative identities. It uses controlled prompt templates and automated analysis to assess sentiment, regard, toxicity, and lexical diversity across completions generated by different models for unmarked, non-queer-marked, and queer-marked subjects.

Additional Metadata
Rights	Academic research only. Please cite the corresponding paper if you use QueerGen in your work.
Organisation	Human-Centered Data Analytics
Citation APA APA Style APA-ALL Style AAA Style Cell Style Chicago Style Harvard Style IEEE Style MLA Style Nature Style Vancouver Style American-Institute-of-Physics Style Council-of-Science-Editors Style BibTex Format Endnote Format RIS Format CSL Format DOIs only Format	Sosto, M. (2025). QueerGen: Evaluating representational harm toward queer identities in language models.

Full Text ( Final Version )

See Also
inProceedings QueerGen: How LLMs reflect societal norms on gender and sexuality in sentence completion tasks M Sosto (Mae), D. S. Martinez Pandiani (Delfina) and L. Hollink (Laura)

QueerGen: Evaluating representational harm toward queer identities in language models

Publication

Publication

inProceedings
QueerGen: How LLMs reflect societal norms on gender and sexuality in sentence completion tasks

Address

CWI researchers

Questions or comments?

QueerGen: Evaluating representational harm toward queer identities in language models

Publication

Publication

inProceedings QueerGen: How LLMs reflect societal norms on gender and sexuality in sentence completion tasks

Workflow

Workflow

Add Content

inProceedings
QueerGen: How LLMs reflect societal norms on gender and sexuality in sentence completion tasks