Towards Multilingual LLM Evaluation for European Languages

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Thellmann, Klaudia, Stadler, Bernhard, Fromm, Michael, Buschhoff, Jasper Schulze, Jude, Alex, Barth, Fabio, Leveling, Johannes, Flores-Herr, Nicolas, Köhler, Joachim, Jäkel, René, Ali, Mehdi
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912075439144960
author Thellmann, Klaudia
Stadler, Bernhard
Fromm, Michael
Buschhoff, Jasper Schulze
Jude, Alex
Barth, Fabio
Leveling, Johannes
Flores-Herr, Nicolas
Köhler, Joachim
Jäkel, René
Ali, Mehdi
author_facet Thellmann, Klaudia
Stadler, Bernhard
Fromm, Michael
Buschhoff, Jasper Schulze
Jude, Alex
Barth, Fabio
Leveling, Johannes
Flores-Herr, Nicolas
Köhler, Joachim
Jäkel, René
Ali, Mehdi
contents The rise of Large Language Models (LLMs) has revolutionized natural language processing across numerous languages and tasks. However, evaluating LLM performance in a consistent and meaningful way across multiple European languages remains challenging, especially due to the scarcity of language-parallel multilingual benchmarks. We introduce a multilingual evaluation approach tailored for European languages. We employ translated versions of five widely-used benchmarks to assess the capabilities of 40 LLMs across 21 European languages. Our contributions include examining the effectiveness of translated benchmarks, assessing the impact of different translation services, and offering a multilingual evaluation framework for LLMs that includes newly created datasets: EU20-MMLU, EU20-HellaSwag, EU20-ARC, EU20-TruthfulQA, and EU20-GSM8K. The benchmarks and results are made publicly available to encourage further research in multilingual LLM evaluation.
format Preprint
id arxiv_https___arxiv_org_abs_2410_08928
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Towards Multilingual LLM Evaluation for European Languages
Thellmann, Klaudia
Stadler, Bernhard
Fromm, Michael
Buschhoff, Jasper Schulze
Jude, Alex
Barth, Fabio
Leveling, Johannes
Flores-Herr, Nicolas
Köhler, Joachim
Jäkel, René
Ali, Mehdi
Computation and Language
Artificial Intelligence
Machine Learning
The rise of Large Language Models (LLMs) has revolutionized natural language processing across numerous languages and tasks. However, evaluating LLM performance in a consistent and meaningful way across multiple European languages remains challenging, especially due to the scarcity of language-parallel multilingual benchmarks. We introduce a multilingual evaluation approach tailored for European languages. We employ translated versions of five widely-used benchmarks to assess the capabilities of 40 LLMs across 21 European languages. Our contributions include examining the effectiveness of translated benchmarks, assessing the impact of different translation services, and offering a multilingual evaluation framework for LLMs that includes newly created datasets: EU20-MMLU, EU20-HellaSwag, EU20-ARC, EU20-TruthfulQA, and EU20-GSM8K. The benchmarks and results are made publicly available to encourage further research in multilingual LLM evaluation.
title Towards Multilingual LLM Evaluation for European Languages
topic Computation and Language
Artificial Intelligence
Machine Learning
url https://arxiv.org/abs/2410.08928