From One to Many: Expanding the Scope of Toxicity Mitigation in Language Models

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Pozzobon, Luiza, Lewis, Patrick, Hooker, Sara, Ermis, Beyza
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866929365284028416
author Pozzobon, Luiza
Lewis, Patrick
Hooker, Sara
Ermis, Beyza
author_facet Pozzobon, Luiza
Lewis, Patrick
Hooker, Sara
Ermis, Beyza
contents To date, toxicity mitigation in language models has almost entirely been focused on single-language settings. As language models embrace multilingual capabilities, it's crucial our safety measures keep pace. Recognizing this research gap, our approach expands the scope of conventional toxicity mitigation to address the complexities presented by multiple languages. In the absence of sufficient annotated datasets across languages, we employ translated data to evaluate and enhance our mitigation techniques. We also compare finetuning mitigation approaches against retrieval-augmented techniques under both static and continual toxicity mitigation scenarios. This allows us to examine the effects of translation quality and the cross-lingual transfer on toxicity mitigation. We also explore how model size and data quantity affect the success of these mitigation efforts. Covering nine languages, our study represents a broad array of linguistic families and levels of resource availability, ranging from high to mid-resource languages. Through comprehensive experiments, we provide insights into the complexities of multilingual toxicity mitigation, offering valuable insights and paving the way for future research in this increasingly important field. Code and data are available at https://github.com/for-ai/goodtriever.
format Preprint
id arxiv_https___arxiv_org_abs_2403_03893
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle From One to Many: Expanding the Scope of Toxicity Mitigation in Language Models
Pozzobon, Luiza
Lewis, Patrick
Hooker, Sara
Ermis, Beyza
Computation and Language
Artificial Intelligence
To date, toxicity mitigation in language models has almost entirely been focused on single-language settings. As language models embrace multilingual capabilities, it's crucial our safety measures keep pace. Recognizing this research gap, our approach expands the scope of conventional toxicity mitigation to address the complexities presented by multiple languages. In the absence of sufficient annotated datasets across languages, we employ translated data to evaluate and enhance our mitigation techniques. We also compare finetuning mitigation approaches against retrieval-augmented techniques under both static and continual toxicity mitigation scenarios. This allows us to examine the effects of translation quality and the cross-lingual transfer on toxicity mitigation. We also explore how model size and data quantity affect the success of these mitigation efforts. Covering nine languages, our study represents a broad array of linguistic families and levels of resource availability, ranging from high to mid-resource languages. Through comprehensive experiments, we provide insights into the complexities of multilingual toxicity mitigation, offering valuable insights and paving the way for future research in this increasingly important field. Code and data are available at https://github.com/for-ai/goodtriever.
title From One to Many: Expanding the Scope of Toxicity Mitigation in Language Models
topic Computation and Language
Artificial Intelligence
url https://arxiv.org/abs/2403.03893