<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="akbar-etal-2026-bayesian">
    <titleInfo>
        <title>Bayesian Belief Layer for Controllable Opinion Dynamics in LLM Agents</title>
    </titleInfo>
    <name type="personal">
        <namePart type="given">Hafsa</namePart>
        <namePart type="family">Akbar</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Daniel</namePart>
        <namePart type="family">Platnick</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Marjan</namePart>
        <namePart type="family">Alirezaie</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Hossein</namePart>
        <namePart type="family">Rahnama</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Alex</namePart>
        <namePart type="family">Pentland</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 Second Workshop for REsearch on Agent Language Models (REALM 2026)</title>
        </titleInfo>
        <name type="personal">
            <namePart type="given">Ehsan</namePart>
            <namePart type="family">Kamalloo</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Nicolas</namePart>
            <namePart type="family">Gontier</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Nikolai</namePart>
            <namePart type="family">Rozanov</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Maryam</namePart>
            <namePart type="family">Hashemzadeh</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Akhil</namePart>
            <namePart type="family">Arora</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">979-8-89176-485-9</identifier>
    </relatedItem>
    <abstract>LLM agents in social simulation revise their opinions implicitly, in context: how open an agent is to persuasion can neither be specified nor verified, and collective outcomes inherit the model’s training prior. We introduce Bayesian Chronicle Agents (BCA), a minimal belief layer separating what an agent believes from how it speaks. Each stance is a probability, updated by one Bayesian step per utterance heard. A single prior-strength parameter ąppa encodes stubbornness, modeled after its role in Friedkin–Johnsen (FJ) opinion dynamics. We then sweep this parameter to yield three canonical regimes of opinion dynamics on demand (consensus, persistent disagreement, committed-minority influence), with persistent disagreement matching the FJ closed-form fixed points at R² = 0.93–0.99. We further show that prescribed ąppa remains recoverable after the language round-trip, with perfect rank-order recovery across all four models. Explicit belief also makes simulation auditable: the layer surfaces systematic per-model stance biases that end-to-end simulation would silently absorb.</abstract>
    <identifier type="citekey">akbar-etal-2026-bayesian</identifier>
    <location>
        <url>https://aclanthology.org/2026.realm-1.55/</url>
    </location>
    <part>
        <date>2026-10</date>
        <extent unit="page">
            <start>807</start>
            <end>815</end>
        </extent>
    </part>
</mods>
</modsCollection>

%0 Conference Proceedings
%T Bayesian Belief Layer for Controllable Opinion Dynamics in LLM Agents
%A Akbar, Hafsa
%A Platnick, Daniel
%A Alirezaie, Marjan
%A Rahnama, Hossein
%A Pentland, Alex
%Y Kamalloo, Ehsan
%Y Gontier, Nicolas
%Y Rozanov, Nikolai
%Y Hashemzadeh, Maryam
%Y Arora, Akhil
%S Proceedings of the Second Workshop for REsearch on Agent Language Models (REALM 2026)
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-485-9
%F akbar-etal-2026-bayesian
%X LLM agents in social simulation revise their opinions implicitly, in context: how open an agent is to persuasion can neither be specified nor verified, and collective outcomes inherit the model’s training prior. We introduce Bayesian Chronicle Agents (BCA), a minimal belief layer separating what an agent believes from how it speaks. Each stance is a probability, updated by one Bayesian step per utterance heard. A single prior-strength parameter ąppa encodes stubbornness, modeled after its role in Friedkin–Johnsen (FJ) opinion dynamics. We then sweep this parameter to yield three canonical regimes of opinion dynamics on demand (consensus, persistent disagreement, committed-minority influence), with persistent disagreement matching the FJ closed-form fixed points at R² = 0.93–0.99. We further show that prescribed ąppa remains recoverable after the language round-trip, with perfect rank-order recovery across all four models. Explicit belief also makes simulation auditable: the layer surfaces systematic per-model stance biases that end-to-end simulation would silently absorb.
%U https://aclanthology.org/2026.realm-1.55/
%P 807-815

Markdown (Informal)

[Bayesian Belief Layer for Controllable Opinion Dynamics in LLM Agents](https://aclanthology.org/2026.realm-1.55/) (Akbar et al., REALM 2026)

• Bayesian Belief Layer for Controllable Opinion Dynamics in LLM Agents (Akbar et al., REALM 2026)

ACL

• Hafsa Akbar, Daniel Platnick, Marjan Alirezaie, Hossein Rahnama, and Alex Pentland. 2026. Bayesian Belief Layer for Controllable Opinion Dynamics in LLM Agents. In Proceedings of the Second Workshop for REsearch on Agent Language Models (REALM 2026), pages 807–815, Budapest, Hungary. Association for Computational Linguistics.