Do Large Language Models Have an English Accent? Evaluating and Improving the Naturalness of Multilingual LLMs

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Guo, Yanzhu, Conia, Simone, Zhou, Zelin, Li, Min, Potdar, Saloni, Xiao, Henry
Formato: Preprint
Publicado: 2024
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866918104486903808
author Guo, Yanzhu
Conia, Simone
Zhou, Zelin
Li, Min
Potdar, Saloni
Xiao, Henry
author_facet Guo, Yanzhu
Conia, Simone
Zhou, Zelin
Li, Min
Potdar, Saloni
Xiao, Henry
contents Current Large Language Models (LLMs) are predominantly designed with English as the primary language, and even the few that are multilingual tend to exhibit strong English-centric biases. Much like speakers who might produce awkward expressions when learning a second language, LLMs often generate unnatural outputs in non-English languages, reflecting English-centric patterns in both vocabulary and grammar. Despite the importance of this issue, the naturalness of multilingual LLM outputs has received limited attention. In this paper, we address this gap by introducing novel automatic corpus-level metrics to assess the lexical and syntactic naturalness of LLM outputs in a multilingual context. Using our new metrics, we evaluate state-of-the-art LLMs on a curated benchmark in French and Chinese, revealing a tendency towards English-influenced patterns. To mitigate this issue, we also propose a simple and effective alignment method to improve the naturalness of an LLM in a target language and domain, achieving consistent improvements in naturalness without compromising the performance on general-purpose benchmarks. Our work highlights the importance of developing multilingual metrics, resources and methods for the new wave of multilingual LLMs.
format Preprint
id arxiv_https___arxiv_org_abs_2410_15956
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Do Large Language Models Have an English Accent? Evaluating and Improving the Naturalness of Multilingual LLMs
Guo, Yanzhu
Conia, Simone
Zhou, Zelin
Li, Min
Potdar, Saloni
Xiao, Henry
Computation and Language
Artificial Intelligence
Current Large Language Models (LLMs) are predominantly designed with English as the primary language, and even the few that are multilingual tend to exhibit strong English-centric biases. Much like speakers who might produce awkward expressions when learning a second language, LLMs often generate unnatural outputs in non-English languages, reflecting English-centric patterns in both vocabulary and grammar. Despite the importance of this issue, the naturalness of multilingual LLM outputs has received limited attention. In this paper, we address this gap by introducing novel automatic corpus-level metrics to assess the lexical and syntactic naturalness of LLM outputs in a multilingual context. Using our new metrics, we evaluate state-of-the-art LLMs on a curated benchmark in French and Chinese, revealing a tendency towards English-influenced patterns. To mitigate this issue, we also propose a simple and effective alignment method to improve the naturalness of an LLM in a target language and domain, achieving consistent improvements in naturalness without compromising the performance on general-purpose benchmarks. Our work highlights the importance of developing multilingual metrics, resources and methods for the new wave of multilingual LLMs.
title Do Large Language Models Have an English Accent? Evaluating and Improving the Naturalness of Multilingual LLMs
topic Computation and Language
Artificial Intelligence
url https://arxiv.org/abs/2410.15956