<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="kang-etal-2026-polar">
    <titleInfo>
        <title>POLAR: Ontology-Guided Risk Prevention for Tool-Calling LLM Agents</title>
    </titleInfo>
    <name type="personal">
        <namePart type="given">Yunju</namePart>
        <namePart type="family">Kang</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Seonghyeon</namePart>
        <namePart type="family">Cho</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Irene</namePart>
        <namePart type="family">Li</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Yeo</namePart>
        <namePart type="given">Chan</namePart>
        <namePart type="family">Yoon</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Chanjun</namePart>
        <namePart type="family">Park</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <originInfo>
        <dateIssued>2026-11</dateIssued>
    </originInfo>
    <typeOfResource>text</typeOfResource>
    <relatedItem type="host">
        <titleInfo>
            <title>Findings of the Association for Computational Linguistics: IJCNLP-AACL 2026</title>
        </titleInfo>
        <name type="personal">
            <namePart type="given">Yue</namePart>
            <namePart type="family">Zhang</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Ruifeng</namePart>
            <namePart type="family">Xu</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Ayu</namePart>
            <namePart type="family">Purwarianti</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Tong</namePart>
            <namePart type="family">Rong</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Jun</namePart>
            <namePart type="family">Suzuki</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <originInfo>
            <publisher>Association for Computational Linguistics</publisher>
            <place>
                <placeTerm type="text">Hengqin, China</placeTerm>
            </place>
        </originInfo>
        <genre authority="marcgt">conference publication</genre>
        <identifier type="isbn">979-8-89176-444-6</identifier>
    </relatedItem>
    <abstract>LLM tool-use agents operate in dynamic environments where many actions carry operational risk. However, most safety mechanisms react only after errors manifest. Existing pre-emptive approaches either fine-tune the agent on chain-of-thought deliberation or compile natural-language guardrails into runtime checks, but they do so without exposing a structural, auditable verdict. We propose POLAR, a guardrail framework for small tool-calling agents that assesses reversibility through a structured two-layer ontology. POLAR assigns each action a graded reversibility score by deriving a candidate inverse sequence; calls failing a threshold are pruned before execution. Evaluated on τ²-bench across six agent models, POLAR improves mean task reward by 0.11 to 0.18 points on airline for four of six agents, but only eight of eighteen model–domain cells improve overall; retail and stronger agents often regress. POLAR provides an auditable structural check and characterizes its task-utility trade-offs. Reward is not a direct measure of prevented harm.</abstract>
    <identifier type="citekey">kang-etal-2026-polar</identifier>
    <location>
        <url>https://aclanthology.org/2026.findings-aacl.220/</url>
    </location>
    <part>
        <date>2026-11</date>
        <extent unit="page">
            <start>4312</start>
            <end>4328</end>
        </extent>
    </part>
</mods>
</modsCollection>

%0 Conference Proceedings
%T POLAR: Ontology-Guided Risk Prevention for Tool-Calling LLM Agents
%A Kang, Yunju
%A Cho, Seonghyeon
%A Li, Irene
%A Yoon, Yeo Chan
%A Park, Chanjun
%Y Zhang, Yue
%Y Xu, Ruifeng
%Y Purwarianti, Ayu
%Y Rong, Tong
%Y Suzuki, Jun
%S Findings of the Association for Computational Linguistics: IJCNLP-AACL 2026
%D 2026
%8 November
%I Association for Computational Linguistics
%C Hengqin, China
%@ 979-8-89176-444-6
%F kang-etal-2026-polar
%X LLM tool-use agents operate in dynamic environments where many actions carry operational risk. However, most safety mechanisms react only after errors manifest. Existing pre-emptive approaches either fine-tune the agent on chain-of-thought deliberation or compile natural-language guardrails into runtime checks, but they do so without exposing a structural, auditable verdict. We propose POLAR, a guardrail framework for small tool-calling agents that assesses reversibility through a structured two-layer ontology. POLAR assigns each action a graded reversibility score by deriving a candidate inverse sequence; calls failing a threshold are pruned before execution. Evaluated on τ²-bench across six agent models, POLAR improves mean task reward by 0.11 to 0.18 points on airline for four of six agents, but only eight of eighteen model–domain cells improve overall; retail and stronger agents often regress. POLAR provides an auditable structural check and characterizes its task-utility trade-offs. Reward is not a direct measure of prevented harm.
%U https://aclanthology.org/2026.findings-aacl.220/
%P 4312-4328

Markdown (Informal)

[POLAR: Ontology-Guided Risk Prevention for Tool-Calling LLM Agents](https://aclanthology.org/2026.findings-aacl.220/) (Kang et al., Findings 2026)

• POLAR: Ontology-Guided Risk Prevention for Tool-Calling LLM Agents (Kang et al., Findings 2026)

ACL

• Yunju Kang, Seonghyeon Cho, Irene Li, Yeo Chan Yoon, and Chanjun Park. 2026. POLAR: Ontology-Guided Risk Prevention for Tool-Calling LLM Agents. In Findings of the Association for Computational Linguistics: IJCNLP-AACL 2026, pages 4312–4328, Hengqin, China. Association for Computational Linguistics.