<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="hu-etal-2026-straight">
    <titleInfo>
        <title>The Straight and Narrow: Do LLMs Possess an Internal Moral Path?</title>
    </titleInfo>
    <name type="personal">
        <namePart type="given">Luoming</namePart>
        <namePart type="family">Hu</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Jingjie</namePart>
        <namePart type="family">Zeng</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Liang</namePart>
        <namePart type="family">Yang</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Hongfei</namePart>
        <namePart type="family">Lin</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>Enhancing the moral alignment of Large Language Models (LLMs) is a critical challenge, yet current techniques like RLHF often function as superficial behavioral guardrails, leaving the model’s internal moral representations untouched and susceptible to complex jailbreaks. In this paper, we propose a shift from behavioral intervention to latent space steering by leveraging Moral Foundations Theory (MFT). To ensure these internal representations capture moral concepts shared across different languages rather than superficial linguistic patterns, we first conduct a cross-lingual linear probing analysis, validating a shared yet culturally nuanced moral subspace between Chinese and English. Establishing this robust semantic foundation allows us to extract steerable Moral Vectors with high causal efficacy at both internal and behavioral levels. Finally, we introduce Adaptive Moral Fusion (AMF), a dynamic inference-time intervention that balances the safety-helpfulness trade-off. Empirical results demonstrate that AMF serves as a targeted intrinsic defense, significantly reducing jailbreak success rates on HarmBench while minimizing incorrect refusals on XSTest compared to standard baselines.</abstract>
    <identifier type="citekey">hu-etal-2026-straight</identifier>
    <location>
        <url>https://aclanthology.org/2026.emnlp-main.1112/</url>
    </location>
    <part>
        <date>2026-10</date>
        <extent unit="page">
            <start>25460</start>
            <end>25482</end>
        </extent>
    </part>
</mods>
</modsCollection>

%0 Conference Proceedings
%T The Straight and Narrow: Do LLMs Possess an Internal Moral Path?
%A Hu, Luoming
%A Zeng, Jingjie
%A Yang, Liang
%A Lin, Hongfei
%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 hu-etal-2026-straight
%X Enhancing the moral alignment of Large Language Models (LLMs) is a critical challenge, yet current techniques like RLHF often function as superficial behavioral guardrails, leaving the model’s internal moral representations untouched and susceptible to complex jailbreaks. In this paper, we propose a shift from behavioral intervention to latent space steering by leveraging Moral Foundations Theory (MFT). To ensure these internal representations capture moral concepts shared across different languages rather than superficial linguistic patterns, we first conduct a cross-lingual linear probing analysis, validating a shared yet culturally nuanced moral subspace between Chinese and English. Establishing this robust semantic foundation allows us to extract steerable Moral Vectors with high causal efficacy at both internal and behavioral levels. Finally, we introduce Adaptive Moral Fusion (AMF), a dynamic inference-time intervention that balances the safety-helpfulness trade-off. Empirical results demonstrate that AMF serves as a targeted intrinsic defense, significantly reducing jailbreak success rates on HarmBench while minimizing incorrect refusals on XSTest compared to standard baselines.
%U https://aclanthology.org/2026.emnlp-main.1112/
%P 25460-25482

Markdown (Informal)

[The Straight and Narrow: Do LLMs Possess an Internal Moral Path?](https://aclanthology.org/2026.emnlp-main.1112/) (Hu et al., EMNLP 2026)

• The Straight and Narrow: Do LLMs Possess an Internal Moral Path? (Hu et al., EMNLP 2026)

ACL

• Luoming Hu, Jingjie Zeng, Liang Yang, and Hongfei Lin. 2026. The Straight and Narrow: Do LLMs Possess an Internal Moral Path?. In Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing, pages 25460–25482, Budapest, Hungary. Association for Computational Linguistics.