<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="belem-etal-2026-may">
    <titleInfo>
        <title>From ‘May’ to ‘Is’: Certainty Distortion in Language Model Rewriting</title>
    </titleInfo>
    <name type="personal">
        <namePart type="given">Catarina</namePart>
        <namePart type="given">G</namePart>
        <namePart type="family">Belém</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Shang</namePart>
        <namePart type="family">Wu</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Randol</namePart>
        <namePart type="given">H</namePart>
        <namePart type="family">Yao</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Mark</namePart>
        <namePart type="family">Steyvers</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Sameer</namePart>
        <namePart type="family">Singh</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Padhraic</namePart>
        <namePart type="family">Smyth</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>Humans increasingly turn to Language Models (LMs) in ways that shape beliefs and drive decisions, including discussing, rewriting, and summarizing information from scientific articles, news, and medical reports. However, in these domains, where it often matters how confidently a claim is expressed, little is known about whether LMs faithfully preserve the degree of confidence. In this work, we investigate _certainty distortion_ in LMs, defined as meaningful changes in expressed certainty during transformations intended to preserve meaning. We propose an LM-based evaluation metric that is consistent with population-level judgments of certainty. Using this metric, we characterize certainty distortion across different sizes and families of models in the context of scientific and medical communication tasks. Our results show that certainty distortion affects up to 75% of LM outputs and is systematically asymmetric in rewriting tasks with most LMs being 1.5-2x more likely to increase the expressed certainty than to decrease it. These effects can compound over repeated paraphrasing: in the medical domain, Claude Haiku (4.5) increases certainty in 20% of examples after a single iteration, increasing to 40% after five iterations. Prompt-based interventions reduce overall certainty distortion but do not eliminate it. Together, these findings reveal a general bias toward inflating expressed certainty, with direct implications for users who rely on LMs in high-stakes domains.</abstract>
    <identifier type="citekey">belem-etal-2026-may</identifier>
    <location>
        <url>https://aclanthology.org/2026.emnlp-main.931/</url>
    </location>
    <part>
        <date>2026-10</date>
        <extent unit="page">
            <start>21426</start>
            <end>21464</end>
        </extent>
    </part>
</mods>
</modsCollection>

%0 Conference Proceedings
%T From ‘May’ to ‘Is’: Certainty Distortion in Language Model Rewriting
%A Belém, Catarina G.
%A Wu, Shang
%A Yao, Randol H.
%A Steyvers, Mark
%A Singh, Sameer
%A Smyth, Padhraic
%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 belem-etal-2026-may
%X Humans increasingly turn to Language Models (LMs) in ways that shape beliefs and drive decisions, including discussing, rewriting, and summarizing information from scientific articles, news, and medical reports. However, in these domains, where it often matters how confidently a claim is expressed, little is known about whether LMs faithfully preserve the degree of confidence. In this work, we investigate _certainty distortion_ in LMs, defined as meaningful changes in expressed certainty during transformations intended to preserve meaning. We propose an LM-based evaluation metric that is consistent with population-level judgments of certainty. Using this metric, we characterize certainty distortion across different sizes and families of models in the context of scientific and medical communication tasks. Our results show that certainty distortion affects up to 75% of LM outputs and is systematically asymmetric in rewriting tasks with most LMs being 1.5-2x more likely to increase the expressed certainty than to decrease it. These effects can compound over repeated paraphrasing: in the medical domain, Claude Haiku (4.5) increases certainty in 20% of examples after a single iteration, increasing to 40% after five iterations. Prompt-based interventions reduce overall certainty distortion but do not eliminate it. Together, these findings reveal a general bias toward inflating expressed certainty, with direct implications for users who rely on LMs in high-stakes domains.
%U https://aclanthology.org/2026.emnlp-main.931/
%P 21426-21464

Markdown (Informal)

[From ’May’ to ’Is’: Certainty Distortion in Language Model Rewriting](https://aclanthology.org/2026.emnlp-main.931/) (Belém et al., EMNLP 2026)

• From ’May’ to ’Is’: Certainty Distortion in Language Model Rewriting (Belém et al., EMNLP 2026)

ACL

• Catarina G Belém, Shang Wu, Randol H. Yao, Mark Steyvers, Sameer Singh, and Padhraic Smyth. 2026. From ’May’ to ’Is’: Certainty Distortion in Language Model Rewriting. In Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing, pages 21426–21464, Budapest, Hungary. Association for Computational Linguistics.