Current decoder-based pre-trained language models (PLMs) successfully demonstrate multilingual capabilities. However, it is unclear how these models handle multilingualism. We analyze the neuron-level internal behavior of multilingual decoder-based PLMs, Specifically examining the existence of neurons that fire ``uniquely for each language'' within decoder-only multilingual PLMs. We analyze six languages: English, German, French, Spanish, Chinese, and Japanese, and show that language-specific neurons are unique, with a slight overlap (< 5%) between languages. These neurons are mainly distributed in the models' first and last few layers. This trend remains consistent across languages and models. Additionally, we tamper with less than 1% of the total neurons in each model during inference and demonstrate that tampering with a few language-specific neurons drastically changes the probability of target language occurrence in text generation.
@article{arxiv.2404.02431,
title = {On the Multilingual Ability of Decoder-based Pre-trained Language Models: Finding and Controlling Language-Specific Neurons},
author = {Takeshi Kojima and Itsuki Okimura and Yusuke Iwasawa and Hitomi Yanaka and Yutaka Matsuo},
journal= {arXiv preprint arXiv:2404.02431},
year = {2024}
}
Comments
Accepted to NAACL2024. Our code is available at https://github.com/kojima-takeshi188/lang_neuron