<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="doan-etal-2026-causal">
    <titleInfo>
        <title>Causal Activation Steering via Sparse Mediation</title>
    </titleInfo>
    <name type="personal">
        <namePart type="given">Toan</namePart>
        <namePart type="family">Doan</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Uyen</namePart>
        <namePart type="family">Le</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Thin</namePart>
        <namePart type="family">Nguyen</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <originInfo>
        <dateIssued>2026-03</dateIssued>
    </originInfo>
    <typeOfResource>text</typeOfResource>
    <relatedItem type="host">
        <titleInfo>
            <title>Findings of the Association for Computational Linguistics: EACL 2026</title>
        </titleInfo>
        <name type="personal">
            <namePart type="given">Vera</namePart>
            <namePart type="family">Demberg</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Kentaro</namePart>
            <namePart type="family">Inui</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Lluís</namePart>
            <namePart type="family">Marquez</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <originInfo>
            <publisher>Association for Computational Linguistics</publisher>
            <place>
                <placeTerm type="text">Rabat, Morocco</placeTerm>
            </place>
        </originInfo>
        <genre authority="marcgt">conference publication</genre>
        <identifier type="isbn">979-8-89176-386-9</identifier>
    </relatedItem>
    <abstract>Activation steering or editing hidden states to control language-model behavior can be framed as a causal mediation problem: inputs induce internal activations, a subset of which act as mediators transmitting targeted behaviors to outputs. We formalize a structural graph over transformer layers and derive front-door—style identification conditions that justify steering through mediating subspaces while preserving non-mediating features, thereby reducing confounding and off-target effects. Within this mediation-first view, we present CAS-BiPO, a sparse mediation steering approach that learns targeted behavioral interventions via regularized training. Empirically, our method achieves 97-100% of dense baseline effectiveness across four behavioral control tasks while using only 10-30% of activation dimensions. Learned masks concentrate 94.3% of steering effects in 26.7% of dimensions, with neurons exhibiting 2.2\times higher activation changes, validating the sparse mediation hypothesis. Our causal framework provides theoretical grounding while CAS-BiPO demonstrates that end-to-end learning of interpretable, reliable interventions is both feasible and advantageous.</abstract>
    <identifier type="citekey">doan-etal-2026-causal</identifier>
    <location>
        <url>https://aclanthology.org/2026.findings-eacl.57/</url>
    </location>
    <part>
        <date>2026-03</date>
        <extent unit="page">
            <start>1079</start>
            <end>1097</end>
        </extent>
    </part>
</mods>
</modsCollection>

%0 Conference Proceedings
%T Causal Activation Steering via Sparse Mediation
%A Doan, Toan
%A Le, Uyen
%A Nguyen, Thin
%Y Demberg, Vera
%Y Inui, Kentaro
%Y Marquez, Lluís
%S Findings of the Association for Computational Linguistics: EACL 2026
%D 2026
%8 March
%I Association for Computational Linguistics
%C Rabat, Morocco
%@ 979-8-89176-386-9
%F doan-etal-2026-causal
%X Activation steering or editing hidden states to control language-model behavior can be framed as a causal mediation problem: inputs induce internal activations, a subset of which act as mediators transmitting targeted behaviors to outputs. We formalize a structural graph over transformer layers and derive front-door—style identification conditions that justify steering through mediating subspaces while preserving non-mediating features, thereby reducing confounding and off-target effects. Within this mediation-first view, we present CAS-BiPO, a sparse mediation steering approach that learns targeted behavioral interventions via regularized training. Empirically, our method achieves 97-100% of dense baseline effectiveness across four behavioral control tasks while using only 10-30% of activation dimensions. Learned masks concentrate 94.3% of steering effects in 26.7% of dimensions, with neurons exhibiting 2.2\times higher activation changes, validating the sparse mediation hypothesis. Our causal framework provides theoretical grounding while CAS-BiPO demonstrates that end-to-end learning of interpretable, reliable interventions is both feasible and advantageous.
%U https://aclanthology.org/2026.findings-eacl.57/
%P 1079-1097

Markdown (Informal)

[Causal Activation Steering via Sparse Mediation](https://aclanthology.org/2026.findings-eacl.57/) (Doan et al., Findings 2026)

• Causal Activation Steering via Sparse Mediation (Doan et al., Findings 2026)

ACL

• Toan Doan, Uyen Le, and Thin Nguyen. 2026. Causal Activation Steering via Sparse Mediation. In Findings of the Association for Computational Linguistics: EACL 2026, pages 1079–1097, Rabat, Morocco. Association for Computational Linguistics.