<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="braun-etal-2026-beyond">
    <titleInfo>
        <title>Beyond Multiple Choice: Evaluating Steering Vectors for Summarization</title>
    </titleInfo>
    <name type="personal">
        <namePart type="given">Joschka</namePart>
        <namePart type="family">Braun</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Carsten</namePart>
        <namePart type="family">Eickhoff</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Seyed</namePart>
        <namePart type="given">Ali</namePart>
        <namePart type="family">Bahrainian</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>Steering vectors are a lightweight method for controlling text properties by adding a learned bias to language model activations at inference time. While predominantly studied for multiple-choice and toy tasks, their effectiveness in free-form generation remains largely unexplored. Moving “Beyond Multiple Choice,” we evaluate steering vectors for controlling topical focus, sentiment, toxicity, and readability in abstractive summaries across the SAMSum, NEWTS, and arXiv datasets. We find that steering effectively controls targeted properties, but high steering strengths consistently induce degenerate repetition and factual hallucinations. Prompting alone preserves summary quality but offers weaker control. Combining both methods yields the strongest control and the most favorable efficacy-quality trade-off at moderate steering strengths. Our work demonstrates that steering vectors face a critical control-quality trade-off in free-form generation, and that hybrid approaches offer best balance in practice.</abstract>
    <identifier type="citekey">braun-etal-2026-beyond</identifier>
    <location>
        <url>https://aclanthology.org/2026.findings-eacl.200/</url>
    </location>
    <part>
        <date>2026-03</date>
        <extent unit="page">
            <start>3849</start>
            <end>3884</end>
        </extent>
    </part>
</mods>
</modsCollection>

%0 Conference Proceedings
%T Beyond Multiple Choice: Evaluating Steering Vectors for Summarization
%A Braun, Joschka
%A Eickhoff, Carsten
%A Bahrainian, Seyed Ali
%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 braun-etal-2026-beyond
%X Steering vectors are a lightweight method for controlling text properties by adding a learned bias to language model activations at inference time. While predominantly studied for multiple-choice and toy tasks, their effectiveness in free-form generation remains largely unexplored. Moving “Beyond Multiple Choice,” we evaluate steering vectors for controlling topical focus, sentiment, toxicity, and readability in abstractive summaries across the SAMSum, NEWTS, and arXiv datasets. We find that steering effectively controls targeted properties, but high steering strengths consistently induce degenerate repetition and factual hallucinations. Prompting alone preserves summary quality but offers weaker control. Combining both methods yields the strongest control and the most favorable efficacy-quality trade-off at moderate steering strengths. Our work demonstrates that steering vectors face a critical control-quality trade-off in free-form generation, and that hybrid approaches offer best balance in practice.
%U https://aclanthology.org/2026.findings-eacl.200/
%P 3849-3884

Markdown (Informal)

[Beyond Multiple Choice: Evaluating Steering Vectors for Summarization](https://aclanthology.org/2026.findings-eacl.200/) (Braun et al., Findings 2026)

• Beyond Multiple Choice: Evaluating Steering Vectors for Summarization (Braun et al., Findings 2026)

ACL

• Joschka Braun, Carsten Eickhoff, and Seyed Ali Bahrainian. 2026. Beyond Multiple Choice: Evaluating Steering Vectors for Summarization. In Findings of the Association for Computational Linguistics: EACL 2026, pages 3849–3884, Rabat, Morocco. Association for Computational Linguistics.