How do Large Language Models Handle Multilingualism?

arXiv:2402.18815v1 Announce Type: new
Abstract: Large language models (LLMs) demonstrate remarkable performance across a spectrum of languages. In this work, we delve into the question: How do LLMs handle multilingualism? We introduce a framework that depicts LLMs' processing of multilingual inputs: In the first several layers, LLMs understand the question, converting multilingual inputs into English to facilitate the task-solving phase. In the intermediate layers, LLMs engage in problem-solving by thinking in English and incorporating multilingual knowledge to obtain factual content, …

abstract arxiv cs.ai cs.cl framework inputs language language models languages large language large language models llms multilingual multilingualism performance processing question spectrum type work