<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="liu-etal-2025-selfelicit">
    <titleInfo>
        <title>SelfElicit: Your Language Model Secretly Knows Where is the Relevant Evidence</title>
    </titleInfo>
    <name type="personal">
        <namePart type="given">Zhining</namePart>
        <namePart type="family">Liu</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Rana</namePart>
        <namePart type="given">Ali</namePart>
        <namePart type="family">Amjad</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Ravinarayana</namePart>
        <namePart type="family">Adkathimar</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Tianxin</namePart>
        <namePart type="family">Wei</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Hanghang</namePart>
        <namePart type="family">Tong</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <originInfo>
        <dateIssued>2025-07</dateIssued>
    </originInfo>
    <typeOfResource>text</typeOfResource>
    <relatedItem type="host">
        <titleInfo>
            <title>Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)</title>
        </titleInfo>
        <name type="personal">
            <namePart type="given">Wanxiang</namePart>
            <namePart type="family">Che</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Joyce</namePart>
            <namePart type="family">Nabende</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Ekaterina</namePart>
            <namePart type="family">Shutova</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Mohammad</namePart>
            <namePart type="given">Taher</namePart>
            <namePart type="family">Pilehvar</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <originInfo>
            <publisher>Association for Computational Linguistics</publisher>
            <place>
                <placeTerm type="text">Vienna, Austria</placeTerm>
            </place>
        </originInfo>
        <genre authority="marcgt">conference publication</genre>
        <identifier type="isbn">979-8-89176-251-0</identifier>
    </relatedItem>
    <abstract>Providing Language Models (LMs) with relevant evidence in the context (either via retrieval or user-provided) can significantly improve their ability to provide better-grounded responses. However, recent studies have found that LMs often struggle to fully comprehend and utilize key evidence from the context, especially when it contains noise and irrelevant information—an issue common in real-world scenarios.To address this, we propose SelfElicit, an inference-time approach that helps LMs focus on key contextual evidence through self-guided explicit highlighting.By leveraging the inherent evidence-finding capabilities of LMs using the attention scores of deeper layers, our method automatically identifies and emphasizes key evidence within the input context, facilitating more accurate and grounded responses without additional training or iterative prompting.We demonstrate that SelfElicit brings consistent and significant improvement on multiple evidence-based QA tasks for various LM families while maintaining computational efficiency.Our code and documentation are available at https://github.com/ZhiningLiu1998/SelfElicit.</abstract>
    <identifier type="citekey">liu-etal-2025-selfelicit</identifier>
    <identifier type="doi">10.18653/v1/2025.acl-long.448</identifier>
    <location>
        <url>https://aclanthology.org/2025.acl-long.448/</url>
    </location>
    <part>
        <date>2025-07</date>
        <extent unit="page">
            <start>9153</start>
            <end>9173</end>
        </extent>
    </part>
</mods>
</modsCollection>

%0 Conference Proceedings
%T SelfElicit: Your Language Model Secretly Knows Where is the Relevant Evidence
%A Liu, Zhining
%A Amjad, Rana Ali
%A Adkathimar, Ravinarayana
%A Wei, Tianxin
%A Tong, Hanghang
%Y Che, Wanxiang
%Y Nabende, Joyce
%Y Shutova, Ekaterina
%Y Pilehvar, Mohammad Taher
%S Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)
%D 2025
%8 July
%I Association for Computational Linguistics
%C Vienna, Austria
%@ 979-8-89176-251-0
%F liu-etal-2025-selfelicit
%X Providing Language Models (LMs) with relevant evidence in the context (either via retrieval or user-provided) can significantly improve their ability to provide better-grounded responses. However, recent studies have found that LMs often struggle to fully comprehend and utilize key evidence from the context, especially when it contains noise and irrelevant information—an issue common in real-world scenarios.To address this, we propose SelfElicit, an inference-time approach that helps LMs focus on key contextual evidence through self-guided explicit highlighting.By leveraging the inherent evidence-finding capabilities of LMs using the attention scores of deeper layers, our method automatically identifies and emphasizes key evidence within the input context, facilitating more accurate and grounded responses without additional training or iterative prompting.We demonstrate that SelfElicit brings consistent and significant improvement on multiple evidence-based QA tasks for various LM families while maintaining computational efficiency.Our code and documentation are available at https://github.com/ZhiningLiu1998/SelfElicit.
%R 10.18653/v1/2025.acl-long.448
%U https://aclanthology.org/2025.acl-long.448/
%U https://doi.org/10.18653/v1/2025.acl-long.448
%P 9153-9173

Markdown (Informal)

[SelfElicit: Your Language Model Secretly Knows Where is the Relevant Evidence](https://aclanthology.org/2025.acl-long.448/) (Liu et al., ACL 2025)

• SelfElicit: Your Language Model Secretly Knows Where is the Relevant Evidence (Liu et al., ACL 2025)

ACL

• Zhining Liu, Rana Ali Amjad, Ravinarayana Adkathimar, Tianxin Wei, and Hanghang Tong. 2025. SelfElicit: Your Language Model Secretly Knows Where is the Relevant Evidence. In Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 9153–9173, Vienna, Austria. Association for Computational Linguistics.