<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="owusu-feldman-2026-anchoring">
    <titleInfo>
        <title>Anchoring Depends on Confidence and Post-Training in Language Models</title>
    </titleInfo>
    <name type="personal">
        <namePart type="given">Hillary</namePart>
        <namePart type="given">N</namePart>
        <namePart type="family">Owusu</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Naomi</namePart>
        <namePart type="given">H</namePart>
        <namePart type="family">Feldman</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <originInfo>
        <dateIssued>2026-07</dateIssued>
    </originInfo>
    <typeOfResource>text</typeOfResource>
    <relatedItem type="host">
        <titleInfo>
            <title>Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers)</title>
        </titleInfo>
        <name type="personal">
            <namePart type="given">Maria</namePart>
            <namePart type="family">Liakata</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Viviane</namePart>
            <namePart type="given">P</namePart>
            <namePart type="family">Moreira</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Jiajun</namePart>
            <namePart type="family">Zhang</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">David</namePart>
            <namePart type="family">Jurgens</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <originInfo>
            <publisher>Association for Computational Linguistics</publisher>
            <place>
                <placeTerm type="text">San Diego, California, United States</placeTerm>
            </place>
        </originInfo>
        <genre authority="marcgt">conference publication</genre>
        <identifier type="isbn">979-8-89176-391-3</identifier>
    </relatedItem>
    <abstract>Anchoring bias causes large language models (LLMs) to shift quantitative judgments in response to irrelevant numerical primes. We analyze this bias as a function of model confidence and accuracy in base, instruction-tuned, and distilled variants of Llama and Qwen models. We find that anchoring susceptibility is negatively correlated with model confidence without regard to accuracy: confidently incorrect models resist anchoring as effectively as accurate ones, provided their internal priors are sufficiently strong. We further show that post-training impacts the strength of this relationship, and that models are more susceptible to high anchors than to low anchors. Our findings suggest anchoring resistance is a structural property of distributional concentration (certainty) rather than knowledge correctness (factual accuracy), with implications for deploying LLMs in numerical reasoning tasks.</abstract>
    <identifier type="citekey">owusu-feldman-2026-anchoring</identifier>
    <location>
        <url>https://aclanthology.org/2026.acl-short.16/</url>
    </location>
    <part>
        <date>2026-07</date>
        <extent unit="page">
            <start>174</start>
            <end>180</end>
        </extent>
    </part>
</mods>
</modsCollection>

%0 Conference Proceedings
%T Anchoring Depends on Confidence and Post-Training in Language Models
%A Owusu, Hillary N.
%A Feldman, Naomi H.
%Y Liakata, Maria
%Y Moreira, Viviane P.
%Y Zhang, Jiajun
%Y Jurgens, David
%S Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers)
%D 2026
%8 July
%I Association for Computational Linguistics
%C San Diego, California, United States
%@ 979-8-89176-391-3
%F owusu-feldman-2026-anchoring
%X Anchoring bias causes large language models (LLMs) to shift quantitative judgments in response to irrelevant numerical primes. We analyze this bias as a function of model confidence and accuracy in base, instruction-tuned, and distilled variants of Llama and Qwen models. We find that anchoring susceptibility is negatively correlated with model confidence without regard to accuracy: confidently incorrect models resist anchoring as effectively as accurate ones, provided their internal priors are sufficiently strong. We further show that post-training impacts the strength of this relationship, and that models are more susceptible to high anchors than to low anchors. Our findings suggest anchoring resistance is a structural property of distributional concentration (certainty) rather than knowledge correctness (factual accuracy), with implications for deploying LLMs in numerical reasoning tasks.
%U https://aclanthology.org/2026.acl-short.16/
%P 174-180

Markdown (Informal)

[Anchoring Depends on Confidence and Post-Training in Language Models](https://aclanthology.org/2026.acl-short.16/) (Owusu & Feldman, ACL 2026)

• Anchoring Depends on Confidence and Post-Training in Language Models (Owusu & Feldman, ACL 2026)

ACL

• Hillary N. Owusu and Naomi H. Feldman. 2026. Anchoring Depends on Confidence and Post-Training in Language Models. In Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers), pages 174–180, San Diego, California, United States. Association for Computational Linguistics.