Ensuring consistent safety across multiple languages remains a significant challenge for large language models (LLMs). We introduce Soteria, a lightweight yet powerful strategy that locates and minimally adjusts the "functional heads" most responsible for harmful content generation in each language. By altering only a fraction of parameters, Soteria drastically reduces policy violations without sacrificing overall model performance, even in low-resource settings. To rigorously evaluate our approach, we also present XThreatBench, a specialized multilingual dataset capturing fine-grained harmful behaviors drawn from real policy guidelines. Experiments with leading open-source LLMs (e.g., Llama, Qwen, Mistral) show that Soteria consistently improves safety metrics across high-, mid-, and low-resource languages. These findings highlight a promising path toward scalable, linguistically attuned, and ethically aligned LLMs worldwide.
@article{arxiv.2502.11244,
title = {Soteria: Language-Specific Functional Parameter Steering for Multilingual Safety Alignment},
author = {Somnath Banerjee and Sayan Layek and Pratyush Chatterjee and Animesh Mukherjee and Rima Hazra},
journal= {arXiv preprint arXiv:2502.11244},
year = {2025}
}