@inproceedings{rigouts-terryn-de-lhoneux-2024-exploratory,
title = "Exploratory Study on the Impact of {E}nglish Bias of Generative Large Language Models in {D}utch and {F}rench",
author = "Rigouts Terryn, Ayla and
de Lhoneux, Miryam",
editor = "Balloccu, Simone and
Belz, Anya and
Huidrom, Rudali and
Reiter, Ehud and
Sedoc, Joao and
Thomson, Craig",
booktitle = "Proceedings of the Fourth Workshop on Human Evaluation of NLP Systems (HumEval) @ LREC-COLING 2024",
month = may,
year = "2024",
address = "Torino, Italia",
publisher = "ELRA and ICCL",
url = "https://aclanthology.org/2024.humeval-1.2",
pages = "12--27",
abstract = "The most widely used LLMs like GPT4 and Llama 2 are trained on large amounts of data, mostly in English but are still able to deal with non-English languages. This English bias leads to lower performance in other languages, especially low-resource ones. This paper studies the linguistic quality of LLMs in two non-English high-resource languages: Dutch and French, with a focus on the influence of English. We first construct a comparable corpus of text generated by humans versus LLMs (GPT-4, Zephyr, and GEITje) in the news domain. We proceed to annotate linguistic issues in the LLM-generated texts, obtaining high inter-annotator agreement, and analyse these annotated issues. We find a substantial influence of English for all models under all conditions: on average, 16{\%} of all annotations of linguistic errors or peculiarities had a clear link to English. Fine-tuning a LLM to a target language (GEITje is fine-tuned on Dutch) reduces the number of linguistic issues and probably also the influence of English. We further find that using a more elaborate prompt leads to linguistically better results than a concise prompt. Finally, increasing the temperature for one of the models leads to lower linguistic quality but does not alter the influence of English.",
}
<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="rigouts-terryn-de-lhoneux-2024-exploratory">
<titleInfo>
<title>Exploratory Study on the Impact of English Bias of Generative Large Language Models in Dutch and French</title>
</titleInfo>
<name type="personal">
<namePart type="given">Ayla</namePart>
<namePart type="family">Rigouts Terryn</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Miryam</namePart>
<namePart type="family">de Lhoneux</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2024-05</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the Fourth Workshop on Human Evaluation of NLP Systems (HumEval) @ LREC-COLING 2024</title>
</titleInfo>
<name type="personal">
<namePart type="given">Simone</namePart>
<namePart type="family">Balloccu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Anya</namePart>
<namePart type="family">Belz</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Rudali</namePart>
<namePart type="family">Huidrom</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ehud</namePart>
<namePart type="family">Reiter</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Joao</namePart>
<namePart type="family">Sedoc</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Craig</namePart>
<namePart type="family">Thomson</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>ELRA and ICCL</publisher>
<place>
<placeTerm type="text">Torino, Italia</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>The most widely used LLMs like GPT4 and Llama 2 are trained on large amounts of data, mostly in English but are still able to deal with non-English languages. This English bias leads to lower performance in other languages, especially low-resource ones. This paper studies the linguistic quality of LLMs in two non-English high-resource languages: Dutch and French, with a focus on the influence of English. We first construct a comparable corpus of text generated by humans versus LLMs (GPT-4, Zephyr, and GEITje) in the news domain. We proceed to annotate linguistic issues in the LLM-generated texts, obtaining high inter-annotator agreement, and analyse these annotated issues. We find a substantial influence of English for all models under all conditions: on average, 16% of all annotations of linguistic errors or peculiarities had a clear link to English. Fine-tuning a LLM to a target language (GEITje is fine-tuned on Dutch) reduces the number of linguistic issues and probably also the influence of English. We further find that using a more elaborate prompt leads to linguistically better results than a concise prompt. Finally, increasing the temperature for one of the models leads to lower linguistic quality but does not alter the influence of English.</abstract>
<identifier type="citekey">rigouts-terryn-de-lhoneux-2024-exploratory</identifier>
<location>
<url>https://aclanthology.org/2024.humeval-1.2</url>
</location>
<part>
<date>2024-05</date>
<extent unit="page">
<start>12</start>
<end>27</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Exploratory Study on the Impact of English Bias of Generative Large Language Models in Dutch and French
%A Rigouts Terryn, Ayla
%A de Lhoneux, Miryam
%Y Balloccu, Simone
%Y Belz, Anya
%Y Huidrom, Rudali
%Y Reiter, Ehud
%Y Sedoc, Joao
%Y Thomson, Craig
%S Proceedings of the Fourth Workshop on Human Evaluation of NLP Systems (HumEval) @ LREC-COLING 2024
%D 2024
%8 May
%I ELRA and ICCL
%C Torino, Italia
%F rigouts-terryn-de-lhoneux-2024-exploratory
%X The most widely used LLMs like GPT4 and Llama 2 are trained on large amounts of data, mostly in English but are still able to deal with non-English languages. This English bias leads to lower performance in other languages, especially low-resource ones. This paper studies the linguistic quality of LLMs in two non-English high-resource languages: Dutch and French, with a focus on the influence of English. We first construct a comparable corpus of text generated by humans versus LLMs (GPT-4, Zephyr, and GEITje) in the news domain. We proceed to annotate linguistic issues in the LLM-generated texts, obtaining high inter-annotator agreement, and analyse these annotated issues. We find a substantial influence of English for all models under all conditions: on average, 16% of all annotations of linguistic errors or peculiarities had a clear link to English. Fine-tuning a LLM to a target language (GEITje is fine-tuned on Dutch) reduces the number of linguistic issues and probably also the influence of English. We further find that using a more elaborate prompt leads to linguistically better results than a concise prompt. Finally, increasing the temperature for one of the models leads to lower linguistic quality but does not alter the influence of English.
%U https://aclanthology.org/2024.humeval-1.2
%P 12-27
Markdown (Informal)
[Exploratory Study on the Impact of English Bias of Generative Large Language Models in Dutch and French](https://aclanthology.org/2024.humeval-1.2) (Rigouts Terryn & de Lhoneux, HumEval-WS 2024)
ACL