How Gender Interacts with Political Values: A Case Study on Czech BERT Models

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Ali, Adnan Al, Libovický, Jindřich
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913274642038784
author Ali, Adnan Al
Libovický, Jindřich
author_facet Ali, Adnan Al
Libovický, Jindřich
contents Neural language models, which reach state-of-the-art results on most natural language processing tasks, are trained on large text corpora that inevitably contain value-burdened content and often capture undesirable biases, which the models reflect. This case study focuses on the political biases of pre-trained encoders in Czech and compares them with a representative value survey. Because Czech is a gendered language, we also measure how the grammatical gender coincides with responses to men and women in the survey. We introduce a novel method for measuring the model's perceived political values. We find that the models do not assign statement probability following value-driven reasoning, and there is no systematic difference between feminine and masculine sentences. We conclude that BERT-sized models do not manifest systematic alignment with political values and that the biases observed in the models are rather due to superficial imitation of training data patterns than systematic value beliefs encoded in the models.
format Preprint
id arxiv_https___arxiv_org_abs_2403_13514
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle How Gender Interacts with Political Values: A Case Study on Czech BERT Models
Ali, Adnan Al
Libovický, Jindřich
Computation and Language
Computers and Society
Neural language models, which reach state-of-the-art results on most natural language processing tasks, are trained on large text corpora that inevitably contain value-burdened content and often capture undesirable biases, which the models reflect. This case study focuses on the political biases of pre-trained encoders in Czech and compares them with a representative value survey. Because Czech is a gendered language, we also measure how the grammatical gender coincides with responses to men and women in the survey. We introduce a novel method for measuring the model's perceived political values. We find that the models do not assign statement probability following value-driven reasoning, and there is no systematic difference between feminine and masculine sentences. We conclude that BERT-sized models do not manifest systematic alignment with political values and that the biases observed in the models are rather due to superficial imitation of training data patterns than systematic value beliefs encoded in the models.
title How Gender Interacts with Political Values: A Case Study on Czech BERT Models
topic Computation and Language
Computers and Society
url https://arxiv.org/abs/2403.13514