<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="li-etal-2026-code">
    <titleInfo>
        <title>Code Agents Don’t Understand Root Cause:Evaluating Code Agents on Linux Kernel Fault Localization</title>
    </titleInfo>
    <name type="personal">
        <namePart type="given">Yuekeng</namePart>
        <namePart type="family">Li</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Ao</namePart>
        <namePart type="family">Xu</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Jie</namePart>
        <namePart type="family">Shi</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Sihang</namePart>
        <namePart type="family">Jiang</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Lipeng</namePart>
        <namePart type="family">Ma</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Bo</namePart>
        <namePart type="family">Xu</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Yanghua</namePart>
        <namePart type="family">Xiao</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>Code agents are increasingly used in debugging workflows, where fault localization (FL) is often evaluated as endpoint retrieval: whether a model names the faulty code location. This view is inadequate for Linux kernel debugging, where failure logs may expose downstream symptoms while the erroneous state is introduced earlier and propagates through various dependences. We introduce KernelPath, a 200-instance Linux 5.10 function-level FL benchmark that pairs each root cause function with a source-grounded state propagation path and a dependence layer. KernelPath evaluates both endpoint localization and ordered diagnostic process recovery. Experiments with SOTA agents show that agents often retrieve plausible candidates but remain unreliable at recovering the diagnostic process, with performance degrading sharply under complex dependence relations. Further diagnostic probes show that agents can use explicit state-transition evidence, yet fail to infer how functions affect the erroneous state, drifting toward neighboring objectives such as repair-site selection, or up stream suspicious function retrieval. These findings reveal an unstable task model for kernel FL in current code agents.</abstract>
    <identifier type="citekey">li-etal-2026-code</identifier>
    <location>
        <url>https://aclanthology.org/2026.emnlp-main.2201/</url>
    </location>
    <part>
        <date>2026-10</date>
        <extent unit="page">
            <start>50033</start>
            <end>50053</end>
        </extent>
    </part>
</mods>
</modsCollection>

%0 Conference Proceedings
%T Code Agents Don’t Understand Root Cause:Evaluating Code Agents on Linux Kernel Fault Localization
%A Li, Yuekeng
%A Xu, Ao
%A Shi, Jie
%A Jiang, Sihang
%A Ma, Lipeng
%A Xu, Bo
%A Xiao, Yanghua
%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 li-etal-2026-code
%X Code agents are increasingly used in debugging workflows, where fault localization (FL) is often evaluated as endpoint retrieval: whether a model names the faulty code location. This view is inadequate for Linux kernel debugging, where failure logs may expose downstream symptoms while the erroneous state is introduced earlier and propagates through various dependences. We introduce KernelPath, a 200-instance Linux 5.10 function-level FL benchmark that pairs each root cause function with a source-grounded state propagation path and a dependence layer. KernelPath evaluates both endpoint localization and ordered diagnostic process recovery. Experiments with SOTA agents show that agents often retrieve plausible candidates but remain unreliable at recovering the diagnostic process, with performance degrading sharply under complex dependence relations. Further diagnostic probes show that agents can use explicit state-transition evidence, yet fail to infer how functions affect the erroneous state, drifting toward neighboring objectives such as repair-site selection, or up stream suspicious function retrieval. These findings reveal an unstable task model for kernel FL in current code agents.
%U https://aclanthology.org/2026.emnlp-main.2201/
%P 50033-50053

Markdown (Informal)

[Code Agents Don’t Understand Root Cause:Evaluating Code Agents on Linux Kernel Fault Localization](https://aclanthology.org/2026.emnlp-main.2201/) (Li et al., EMNLP 2026)

• Code Agents Don’t Understand Root Cause:Evaluating Code Agents on Linux Kernel Fault Localization (Li et al., EMNLP 2026)

ACL

• Yuekeng Li, Ao Xu, Jie Shi, Sihang Jiang, Lipeng Ma, Bo Xu, and Yanghua Xiao. 2026. Code Agents Don’t Understand Root Cause:Evaluating Code Agents on Linux Kernel Fault Localization. In Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing, pages 50033–50053, Budapest, Hungary. Association for Computational Linguistics.