<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="vogels-etal-2026-distribution">
    <titleInfo>
        <title>In-Distribution Steering: Balancing Control and Coherence in Language Model Generation</title>
    </titleInfo>
    <name type="personal">
        <namePart type="given">Arthur</namePart>
        <namePart type="family">Vogels</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Benjamin</namePart>
        <namePart type="family">Wong</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Yann</namePart>
        <namePart type="family">Choho</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Annabelle</namePart>
        <namePart type="family">Blangero</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Milan</namePart>
        <namePart type="family">Bhan</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <originInfo>
        <dateIssued>2026-05</dateIssued>
    </originInfo>
    <typeOfResource>text</typeOfResource>
    <relatedItem type="host">
        <titleInfo>
            <title>Proceedings of the Fifteenth Language Resources and Evaluation Conference</title>
        </titleInfo>
        <name type="personal">
            <namePart type="given">Stelios</namePart>
            <namePart type="family">Piperidis</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Núria</namePart>
            <namePart type="family">Bel</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Henk</namePart>
            <namePart type="family">van den Heuvel</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Nancy</namePart>
            <namePart type="family">Ide</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Simon</namePart>
            <namePart type="family">Krek</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Antonio</namePart>
            <namePart type="family">Toral</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <originInfo>
            <publisher>ELRA Language Resource Association</publisher>
            <place>
                <placeTerm type="text">Palma de Mallorca, Spain</placeTerm>
            </place>
        </originInfo>
        <genre authority="marcgt">conference publication</genre>
    </relatedItem>
    <abstract>Activation steering methods control large language model (LLM) behavior by modifying internal activations at inference time. However, most existing activation steering methods rely on a fixed steering strength, leading to either insufficient control or unadapted intervention that degrades text plausibility and coherence. We introduce In-Distribution Steering (IDS), a novel method that adapts steering strength based on the input data distribution in representation space. IDS dynamically adjusts interventions according to how far a given input lies within the distribution, enabling adaptive intervention and generation stability during text generation. Experiments demonstrate that IDS achieves strong accuracy on classification tasks while producing coherent text without collapse, making IDS particularly well suited for real-world applications.</abstract>
    <identifier type="citekey">vogels-etal-2026-distribution</identifier>
    <identifier type="doi">10.63317/4629fxavjicu</identifier>
    <location>
        <url>https://aclanthology.org/2026.lrec-1.163/</url>
    </location>
    <part>
        <date>2026-05</date>
        <extent unit="page">
            <start>2076</start>
            <end>2089</end>
        </extent>
    </part>
</mods>
</modsCollection>

%0 Conference Proceedings
%T In-Distribution Steering: Balancing Control and Coherence in Language Model Generation
%A Vogels, Arthur
%A Wong, Benjamin
%A Choho, Yann
%A Blangero, Annabelle
%A Bhan, Milan
%Y Piperidis, Stelios
%Y Bel, Núria
%Y van den Heuvel, Henk
%Y Ide, Nancy
%Y Krek, Simon
%Y Toral, Antonio
%S Proceedings of the Fifteenth Language Resources and Evaluation Conference
%D 2026
%8 May
%I ELRA Language Resource Association
%C Palma de Mallorca, Spain
%F vogels-etal-2026-distribution
%X Activation steering methods control large language model (LLM) behavior by modifying internal activations at inference time. However, most existing activation steering methods rely on a fixed steering strength, leading to either insufficient control or unadapted intervention that degrades text plausibility and coherence. We introduce In-Distribution Steering (IDS), a novel method that adapts steering strength based on the input data distribution in representation space. IDS dynamically adjusts interventions according to how far a given input lies within the distribution, enabling adaptive intervention and generation stability during text generation. Experiments demonstrate that IDS achieves strong accuracy on classification tasks while producing coherent text without collapse, making IDS particularly well suited for real-world applications.
%R 10.63317/4629fxavjicu
%U https://aclanthology.org/2026.lrec-1.163/
%U https://doi.org/10.63317/4629fxavjicu
%P 2076-2089

Markdown (Informal)

[In-Distribution Steering: Balancing Control and Coherence in Language Model Generation](https://aclanthology.org/2026.lrec-1.163/) (Vogels et al., LREC 2026)

• In-Distribution Steering: Balancing Control and Coherence in Language Model Generation (Vogels et al., LREC 2026)

ACL

• Arthur Vogels, Benjamin Wong, Yann Choho, Annabelle Blangero, and Milan Bhan. 2026. In-Distribution Steering: Balancing Control and Coherence in Language Model Generation. In Proceedings of the Fifteenth Language Resources and Evaluation Conference, pages 2076–2089, Palma de Mallorca, Spain. ELRA Language Resource Association.