Do Multilingual LLMs Think In English?

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Schut, Lisa, Gal, Yarin, Farquhar, Sebastian
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917932485836800
author Schut, Lisa
Gal, Yarin
Farquhar, Sebastian
author_facet Schut, Lisa
Gal, Yarin
Farquhar, Sebastian
contents Large language models (LLMs) have multilingual capabilities and can solve tasks across various languages. However, we show that current LLMs make key decisions in a representation space closest to English, regardless of their input and output languages. Exploring the internal representations with a logit lens for sentences in French, German, Dutch, and Mandarin, we show that the LLM first emits representations close to English for semantically-loaded words before translating them into the target language. We further show that activation steering in these LLMs is more effective when the steering vectors are computed in English rather than in the language of the inputs and outputs. This suggests that multilingual LLMs perform key reasoning steps in a representation that is heavily shaped by English in a way that is not transparent to system users.
format Preprint
id arxiv_https___arxiv_org_abs_2502_15603
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Do Multilingual LLMs Think In English?
Schut, Lisa
Gal, Yarin
Farquhar, Sebastian
Computation and Language
Artificial Intelligence
Machine Learning
Large language models (LLMs) have multilingual capabilities and can solve tasks across various languages. However, we show that current LLMs make key decisions in a representation space closest to English, regardless of their input and output languages. Exploring the internal representations with a logit lens for sentences in French, German, Dutch, and Mandarin, we show that the LLM first emits representations close to English for semantically-loaded words before translating them into the target language. We further show that activation steering in these LLMs is more effective when the steering vectors are computed in English rather than in the language of the inputs and outputs. This suggests that multilingual LLMs perform key reasoning steps in a representation that is heavily shaped by English in a way that is not transparent to system users.
title Do Multilingual LLMs Think In English?
topic Computation and Language
Artificial Intelligence
Machine Learning
url https://arxiv.org/abs/2502.15603