Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2410.03996 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866917795605774336 |
|---|---|
| author | Sancheti, Abhilasha An, Haozhe Rudinger, Rachel |
| author_facet | Sancheti, Abhilasha An, Haozhe Rudinger, Rachel |
| contents | We study the presence of heteronormative biases and prejudice against interracial romantic relationships in large language models by performing controlled name-replacement experiments for the task of relationship prediction. We show that models are less likely to predict romantic relationships for (a) same-gender character pairs than different-gender pairs; and (b) intra/inter-racial character pairs involving Asian names as compared to Black, Hispanic, or White names. We examine the contextualized embeddings of first names and find that gender for Asian names is less discernible than non-Asian names. We discuss the social implications of our findings, underlining the need to prioritize the development of inclusive and equitable technology. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2410_03996 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | On the Influence of Gender and Race in Romantic Relationship Prediction from Large Language Models Sancheti, Abhilasha An, Haozhe Rudinger, Rachel Computation and Language We study the presence of heteronormative biases and prejudice against interracial romantic relationships in large language models by performing controlled name-replacement experiments for the task of relationship prediction. We show that models are less likely to predict romantic relationships for (a) same-gender character pairs than different-gender pairs; and (b) intra/inter-racial character pairs involving Asian names as compared to Black, Hispanic, or White names. We examine the contextualized embeddings of first names and find that gender for Asian names is less discernible than non-Asian names. We discuss the social implications of our findings, underlining the need to prioritize the development of inclusive and equitable technology. |
| title | On the Influence of Gender and Race in Romantic Relationship Prediction from Large Language Models |
| topic | Computation and Language |
| url | https://arxiv.org/abs/2410.03996 |