<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="su-etal-2026-contrastive">
    <titleInfo>
        <title>Contrastive Neuron Regulation for Chain-of-Thought Reasoning</title>
    </titleInfo>
    <name type="personal">
        <namePart type="given">Zitao</namePart>
        <namePart type="family">Su</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Xinyu</namePart>
        <namePart type="family">Tang</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Xin</namePart>
        <namePart type="family">Zhao</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>Chain-of-thought (CoT) reasoning is commonly improved through external mechanisms such as prompting, sampling, and decoding, while the internal units shaping reasoning trajectories remain less understood. We study CoT from the perspective of reasoning trajectory structure and find that incorrect trajectories are more redundant and repetitive, whereas correct ones are more lexically diverse and information-dense. Motivated by this observation, we propose Contrastive Neuron Regulation (CNR), a sparse, backbone-frozen framework for neuron-level regulation of CoT reasoning. CNR constructs correctness-controlled expanded–concise pairs of correct trajectories, identifies MLP neurons associated with the expansion contrast, and sparsely suppresses them during generation. We consider both a fixed-coefficient inference-time intervention and an adaptive variant that learns neuron-specific suppression coefficients on a subject-specific development split while keeping the backbone frozen. Across two reasoning models and seven MATH subjects, CNR improves macro-average accuracy over the no-intervention baseline by 2.6 and 3.2 percentage points, respectively, while intervening on less than 1% of MLP neurons. The code is available at https://github.com/RUCAIBox/Contrastive-Neuron-Regulation.</abstract>
    <identifier type="citekey">su-etal-2026-contrastive</identifier>
    <location>
        <url>https://aclanthology.org/2026.emnlp-main.1846/</url>
    </location>
    <part>
        <date>2026-10</date>
        <extent unit="page">
            <start>41986</start>
            <end>42005</end>
        </extent>
    </part>
</mods>
</modsCollection>

%0 Conference Proceedings
%T Contrastive Neuron Regulation for Chain-of-Thought Reasoning
%A Su, Zitao
%A Tang, Xinyu
%A Zhao, Xin
%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 su-etal-2026-contrastive
%X Chain-of-thought (CoT) reasoning is commonly improved through external mechanisms such as prompting, sampling, and decoding, while the internal units shaping reasoning trajectories remain less understood. We study CoT from the perspective of reasoning trajectory structure and find that incorrect trajectories are more redundant and repetitive, whereas correct ones are more lexically diverse and information-dense. Motivated by this observation, we propose Contrastive Neuron Regulation (CNR), a sparse, backbone-frozen framework for neuron-level regulation of CoT reasoning. CNR constructs correctness-controlled expanded–concise pairs of correct trajectories, identifies MLP neurons associated with the expansion contrast, and sparsely suppresses them during generation. We consider both a fixed-coefficient inference-time intervention and an adaptive variant that learns neuron-specific suppression coefficients on a subject-specific development split while keeping the backbone frozen. Across two reasoning models and seven MATH subjects, CNR improves macro-average accuracy over the no-intervention baseline by 2.6 and 3.2 percentage points, respectively, while intervening on less than 1% of MLP neurons. The code is available at https://github.com/RUCAIBox/Contrastive-Neuron-Regulation.
%U https://aclanthology.org/2026.emnlp-main.1846/
%P 41986-42005

Markdown (Informal)

[Contrastive Neuron Regulation for Chain-of-Thought Reasoning](https://aclanthology.org/2026.emnlp-main.1846/) (Su et al., EMNLP 2026)

• Contrastive Neuron Regulation for Chain-of-Thought Reasoning (Su et al., EMNLP 2026)

ACL

• Zitao Su, Xinyu Tang, and Xin Zhao. 2026. Contrastive Neuron Regulation for Chain-of-Thought Reasoning. In Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing, pages 41986–42005, Budapest, Hungary. Association for Computational Linguistics.