Preference Tuning For Toxicity Mitigation Generalizes Across Languages

X Li, ZX Yong, SH Bach - arXiv preprint arXiv:2406.16235, 2024 - arxiv.org
Detoxifying multilingual Large Language Models (LLMs) has become crucial due to their
increasing global use. In this work, we explore zero-shot cross-lingual generalization of …