Beyond Binary Gender Labels

The official repository for Beyond Binary Gender Labels: Revealing Gender Bias in LLMs through Gender-Neutral Name Predictions

Name-based gender prediction has traditionally categorized individuals as either female or male based on their names, using a binary classification system. That binary approach can be problematic in the cases of gender-neutral names that do not align with any one gender, among other reasons. Relying solely on binary gender categories without recognizing gender-neutral names can reduce the inclusiveness of gender prediction tasks. We introduce an additional gender category, i.e., "neutral", to study and address potential gender biases in Large Language Models (LLMs). We evaluate the performance of several foundational and large language models in predicting gender based on first names only. Additionally, we investigate the impact of adding birth years to enhance the accuracy of gender prediction, accounting for shifting associations between names and genders over time. Our findings indicate that most LLMs identify male and female names with high accuracy (over 80%) but struggle with gender-neutral names (under 40%), and the accuracy of gender prediction is higher for English-based first names than non-English names. The experimental results show that incorporating the birth year does not improve the overall accuracy of gender prediction, especially for names with evolving gender associations. We recommend using caution when applying LLMs for gender identification in downstream tasks, particularly when dealing with non-binary gender labels.

Datasets

Check the data folder for all datasets we used in this paper, including balanced datasets and dynamic gender label datasets.

Environment

Please set up the environments as follows:

conda create -n <env name> python=3.8.12
pip install transformers pandas openai numpy sentence-transformers datasets

Citation

Please cite the below paper if you intent to use the code for your research.

@inproceedings{you-etal-2024-beyond,
    title = "Beyond Binary Gender Labels: Revealing Gender Bias in {LLM}s through Gender-Neutral Name Predictions",
    author = "You, Zhiwen  and
      Lee, HaeJin  and
      Mishra, Shubhanshu  and
      Jeoung, Sullam  and
      Mishra, Apratim  and
      Kim, Jinseok  and
      Diesner, Jana",
    editor = "Fale{\'n}ska, Agnieszka  and
      Basta, Christine  and
      Costa-juss{\`a}, Marta  and
      Goldfarb-Tarrant, Seraphina  and
      Nozza, Debora",
    booktitle = "Proceedings of the 5th Workshop on Gender Bias in Natural Language Processing (GeBNLP)",
    month = aug,
    year = "2024",
    address = "Bangkok, Thailand",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2024.gebnlp-1.16",
    doi = "10.18653/v1/2024.gebnlp-1.16",
    pages = "255--268",
}