@inproceedings{hou-etal-2026-ai-reviewers,
title = "Do {AI} Reviewers Converge? Diversity Collapse in Model-Generated Peer Review",
author = "Hou, Xuzhe and
Zhang, Shengxin and
Stacy, Riley and
Black, Emily and
Chandrasekaran, Varun",
editor = "Lee, Hung-Yi and
Dev, Sunipa and
Li, Jessy and
Elliott, Desmond",
booktitle = "Findings of the {A}ssociation for {C}omputational {L}inguistics: {EMNLP} 2026",
month = oct,
year = "2026",
address = "Budapest, Hungary",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.findings-emnlp.1035/",
pages = "21856--21885",
ISBN = "979-8-89176-462-0",
abstract = "Large language models (LLMs) are increasingly used to generate or assist reviews. This raises the risk of generative monoculture, where models converge on homogeneous judgments and language patterns, limiting the novelty of scientific progress. We study this phenomenon over 1,943 human reviews from 509 ICLR papers with 10,172 reviews generated by five state-of-the-art LLMs under multiple prompting strategies. We evaluate human and LLM review diversity across several dimensions, including praise-to-critique balance and semantic similarity. Across metrics, LLM-generated reviews exhibit systematic convergence: for example, they are substantially more praise-heavy, and more semantically similar across unrelated papers, suggesting a lack of paper-specific feedback. Specialized review agents reduce{---}but do not eliminate{---}these effects. Our work demonstrates the risk of introducing epistemic monoculture into peer review and develops methods to measure it. If left unchecked, this monoculture could narrow the range of perspectives that shape scientific evaluation."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="hou-etal-2026-ai-reviewers">
<titleInfo>
<title>Do AI Reviewers Converge? Diversity Collapse in Model-Generated Peer Review</title>
</titleInfo>
<name type="personal">
<namePart type="given">Xuzhe</namePart>
<namePart type="family">Hou</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Shengxin</namePart>
<namePart type="family">Zhang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Riley</namePart>
<namePart type="family">Stacy</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Emily</namePart>
<namePart type="family">Black</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Varun</namePart>
<namePart type="family">Chandrasekaran</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>Findings of the Association for Computational Linguistics: EMNLP 2026</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">979-8-89176-462-0</identifier>
</relatedItem>
<abstract>Large language models (LLMs) are increasingly used to generate or assist reviews. This raises the risk of generative monoculture, where models converge on homogeneous judgments and language patterns, limiting the novelty of scientific progress. We study this phenomenon over 1,943 human reviews from 509 ICLR papers with 10,172 reviews generated by five state-of-the-art LLMs under multiple prompting strategies. We evaluate human and LLM review diversity across several dimensions, including praise-to-critique balance and semantic similarity. Across metrics, LLM-generated reviews exhibit systematic convergence: for example, they are substantially more praise-heavy, and more semantically similar across unrelated papers, suggesting a lack of paper-specific feedback. Specialized review agents reduce—but do not eliminate—these effects. Our work demonstrates the risk of introducing epistemic monoculture into peer review and develops methods to measure it. If left unchecked, this monoculture could narrow the range of perspectives that shape scientific evaluation.</abstract>
<identifier type="citekey">hou-etal-2026-ai-reviewers</identifier>
<location>
<url>https://aclanthology.org/2026.findings-emnlp.1035/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>21856</start>
<end>21885</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Do AI Reviewers Converge? Diversity Collapse in Model-Generated Peer Review
%A Hou, Xuzhe
%A Zhang, Shengxin
%A Stacy, Riley
%A Black, Emily
%A Chandrasekaran, Varun
%Y Lee, Hung-Yi
%Y Dev, Sunipa
%Y Li, Jessy
%Y Elliott, Desmond
%S Findings of the Association for Computational Linguistics: EMNLP 2026
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-462-0
%F hou-etal-2026-ai-reviewers
%X Large language models (LLMs) are increasingly used to generate or assist reviews. This raises the risk of generative monoculture, where models converge on homogeneous judgments and language patterns, limiting the novelty of scientific progress. We study this phenomenon over 1,943 human reviews from 509 ICLR papers with 10,172 reviews generated by five state-of-the-art LLMs under multiple prompting strategies. We evaluate human and LLM review diversity across several dimensions, including praise-to-critique balance and semantic similarity. Across metrics, LLM-generated reviews exhibit systematic convergence: for example, they are substantially more praise-heavy, and more semantically similar across unrelated papers, suggesting a lack of paper-specific feedback. Specialized review agents reduce—but do not eliminate—these effects. Our work demonstrates the risk of introducing epistemic monoculture into peer review and develops methods to measure it. If left unchecked, this monoculture could narrow the range of perspectives that shape scientific evaluation.
%U https://aclanthology.org/2026.findings-emnlp.1035/
%P 21856-21885
Markdown (Informal)
[Do AI Reviewers Converge? Diversity Collapse in Model-Generated Peer Review](https://aclanthology.org/2026.findings-emnlp.1035/) (Hou et al., Findings 2026)
ACL