@inproceedings{wright-etal-2026-whose,
title = "What and Whose Knowledge? Measuring Epistemic Diversity in Large Language Models",
author = "Wright, Dustin and
Masud, Sarah and
Moore, Jared and
Yadav, Srishti and
Antoniak, Maria and
Christensen, Peter Ebert and
Park, Chan Young and
Augenstein, Isabelle",
editor = "Lee, Hung-Yi and
Dev, Sunipa and
Li, Jessy and
Elliott, Desmond",
booktitle = "Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing",
month = oct,
year = "2026",
address = "Budapest, Hungary",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.emnlp-main.1542/",
pages = "35186--35203",
ISBN = "979-8-89176-459-0",
abstract = "Large language models (LLMs) are increasingly used as primary knowledge sources, yet their epistemic diversity {---} defined as the diversity of real-world claims in their outputs {---} has never been measured. The dominant paradigm is that overall LLM diversity is low, but this is always with respect to a single point in time, with no reference baseline or consideration for variation across countries. We address this by performing the first systematic study of epistemic diversity in LLMs across time and cultural context, testing 27 LLMs on 155 topics covering 12 countries, resulting in 1.7M responses and 70M individual claims. We find that epistemic diversity has increased substantially over the past three years, a positive counter to recent diversity pessimism. However, despite progress, we find that every LLM is less diverse than a search baseline. The gap is not uniform: RAG can improve diversity, while larger models are counterintuitively less diverse than smaller ones. Moreover, LLMs systematically reflect English-language knowledge over local-language knowledge for country specific topics. Together, these results demonstrate that while progress on epistemic diversity is tangible, it is insufficient and unevenly distributed."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="wright-etal-2026-whose">
<titleInfo>
<title>What and Whose Knowledge? Measuring Epistemic Diversity in Large Language Models</title>
</titleInfo>
<name type="personal">
<namePart type="given">Dustin</namePart>
<namePart type="family">Wright</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Sarah</namePart>
<namePart type="family">Masud</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jared</namePart>
<namePart type="family">Moore</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Srishti</namePart>
<namePart type="family">Yadav</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Maria</namePart>
<namePart type="family">Antoniak</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Peter</namePart>
<namePart type="given">Ebert</namePart>
<namePart type="family">Christensen</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Chan</namePart>
<namePart type="given">Young</namePart>
<namePart type="family">Park</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Isabelle</namePart>
<namePart type="family">Augenstein</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-10</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing</title>
</titleInfo>
<name type="personal">
<namePart type="given">Hung-Yi</namePart>
<namePart type="family">Lee</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Sunipa</namePart>
<namePart type="family">Dev</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jessy</namePart>
<namePart type="family">Li</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Desmond</namePart>
<namePart type="family">Elliott</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Budapest, Hungary</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-459-0</identifier>
</relatedItem>
<abstract>Large language models (LLMs) are increasingly used as primary knowledge sources, yet their epistemic diversity — defined as the diversity of real-world claims in their outputs — has never been measured. The dominant paradigm is that overall LLM diversity is low, but this is always with respect to a single point in time, with no reference baseline or consideration for variation across countries. We address this by performing the first systematic study of epistemic diversity in LLMs across time and cultural context, testing 27 LLMs on 155 topics covering 12 countries, resulting in 1.7M responses and 70M individual claims. We find that epistemic diversity has increased substantially over the past three years, a positive counter to recent diversity pessimism. However, despite progress, we find that every LLM is less diverse than a search baseline. The gap is not uniform: RAG can improve diversity, while larger models are counterintuitively less diverse than smaller ones. Moreover, LLMs systematically reflect English-language knowledge over local-language knowledge for country specific topics. Together, these results demonstrate that while progress on epistemic diversity is tangible, it is insufficient and unevenly distributed.</abstract>
<identifier type="citekey">wright-etal-2026-whose</identifier>
<location>
<url>https://aclanthology.org/2026.emnlp-main.1542/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>35186</start>
<end>35203</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T What and Whose Knowledge? Measuring Epistemic Diversity in Large Language Models
%A Wright, Dustin
%A Masud, Sarah
%A Moore, Jared
%A Yadav, Srishti
%A Antoniak, Maria
%A Christensen, Peter Ebert
%A Park, Chan Young
%A Augenstein, Isabelle
%Y Lee, Hung-Yi
%Y Dev, Sunipa
%Y Li, Jessy
%Y Elliott, Desmond
%S Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-459-0
%F wright-etal-2026-whose
%X Large language models (LLMs) are increasingly used as primary knowledge sources, yet their epistemic diversity — defined as the diversity of real-world claims in their outputs — has never been measured. The dominant paradigm is that overall LLM diversity is low, but this is always with respect to a single point in time, with no reference baseline or consideration for variation across countries. We address this by performing the first systematic study of epistemic diversity in LLMs across time and cultural context, testing 27 LLMs on 155 topics covering 12 countries, resulting in 1.7M responses and 70M individual claims. We find that epistemic diversity has increased substantially over the past three years, a positive counter to recent diversity pessimism. However, despite progress, we find that every LLM is less diverse than a search baseline. The gap is not uniform: RAG can improve diversity, while larger models are counterintuitively less diverse than smaller ones. Moreover, LLMs systematically reflect English-language knowledge over local-language knowledge for country specific topics. Together, these results demonstrate that while progress on epistemic diversity is tangible, it is insufficient and unevenly distributed.
%U https://aclanthology.org/2026.emnlp-main.1542/
%P 35186-35203
Markdown (Informal)
[What and Whose Knowledge? Measuring Epistemic Diversity in Large Language Models](https://aclanthology.org/2026.emnlp-main.1542/) (Wright et al., EMNLP 2026)
ACL
- Dustin Wright, Sarah Masud, Jared Moore, Srishti Yadav, Maria Antoniak, Peter Ebert Christensen, Chan Young Park, and Isabelle Augenstein. 2026. What and Whose Knowledge? Measuring Epistemic Diversity in Large Language Models. In Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing, pages 35186–35203, Budapest, Hungary. Association for Computational Linguistics.