<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="li-etal-2026-aura">
    <titleInfo>
        <title>AURA: Intent-Directed Probing for Implicit-Need Surfacing in Situated LLM Agents</title>
    </titleInfo>
    <name type="personal">
        <namePart type="given">Yang</namePart>
        <namePart type="family">Li</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Jiaxiang</namePart>
        <namePart type="family">Liu</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Jiang</namePart>
        <namePart type="family">Cai</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Mingkun</namePart>
        <namePart type="family">Xu</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>A situated query like “where is Lin Wei?” often encodes more than its literal content: the user may also want to know whether Lin Wei is free, in a good mood, or worth interrupting now. Standard tool-use agents answer the literal question and stop. AURA inserts an inference step between scene perception and tool use that produces an IntentFrame: a structured estimate of the implicit need with a scalar gap score that controls per-query probe budget and tool selection. On a 100-query four-scene implicit-intent benchmark, AURA improves implicit-need coverage over ReAct-style probing (Delta = +0.07, p \ensuremath&lt; 10⌃-6); three of four scenes are individually significant, the gain reproduces on a second backbone, and a prompt ablation attributes the lift to gap calibration rather than answer memorisation. A component ablation localises the gain further: the gap-derived budget ceiling alone outperforms the full architecture (p = 0.02), and bin-wise calibration analysis (r = -0.33) identifies the gap score as a probe trigger rather than a calibrated value estimator. On factual lookup the controller trades raw accuracy for 82% fewer probes and zero forbidden-tool violations on a privacy-sensitive slice; scope conditions are detailed in Limitations. Code, simulator, and benchmark are released at https://github.com/innovation64/AURA, with the implicit-intent benchmark also on the HuggingFace Hub (innovation64/aura-implicit-intent).</abstract>
    <identifier type="citekey">li-etal-2026-aura</identifier>
    <location>
        <url>https://aclanthology.org/2026.emnlp-main.1432/</url>
    </location>
    <part>
        <date>2026-10</date>
        <extent unit="page">
            <start>32700</start>
            <end>32730</end>
        </extent>
    </part>
</mods>
</modsCollection>

%0 Conference Proceedings
%T AURA: Intent-Directed Probing for Implicit-Need Surfacing in Situated LLM Agents
%A Li, Yang
%A Liu, Jiaxiang
%A Cai, Jiang
%A Xu, Mingkun
%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-aura
%X A situated query like “where is Lin Wei?” often encodes more than its literal content: the user may also want to know whether Lin Wei is free, in a good mood, or worth interrupting now. Standard tool-use agents answer the literal question and stop. AURA inserts an inference step between scene perception and tool use that produces an IntentFrame: a structured estimate of the implicit need with a scalar gap score that controls per-query probe budget and tool selection. On a 100-query four-scene implicit-intent benchmark, AURA improves implicit-need coverage over ReAct-style probing (Delta = +0.07, p \ensuremath< 10⌃-6); three of four scenes are individually significant, the gain reproduces on a second backbone, and a prompt ablation attributes the lift to gap calibration rather than answer memorisation. A component ablation localises the gain further: the gap-derived budget ceiling alone outperforms the full architecture (p = 0.02), and bin-wise calibration analysis (r = -0.33) identifies the gap score as a probe trigger rather than a calibrated value estimator. On factual lookup the controller trades raw accuracy for 82% fewer probes and zero forbidden-tool violations on a privacy-sensitive slice; scope conditions are detailed in Limitations. Code, simulator, and benchmark are released at https://github.com/innovation64/AURA, with the implicit-intent benchmark also on the HuggingFace Hub (innovation64/aura-implicit-intent).
%U https://aclanthology.org/2026.emnlp-main.1432/
%P 32700-32730

Markdown (Informal)

[AURA: Intent-Directed Probing for Implicit-Need Surfacing in Situated LLM Agents](https://aclanthology.org/2026.emnlp-main.1432/) (Li et al., EMNLP 2026)

• AURA: Intent-Directed Probing for Implicit-Need Surfacing in Situated LLM Agents (Li et al., EMNLP 2026)

ACL

• Yang Li, Jiaxiang Liu, Jiang Cai, and Mingkun Xu. 2026. AURA: Intent-Directed Probing for Implicit-Need Surfacing in Situated LLM Agents. In Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing, pages 32700–32730, Budapest, Hungary. Association for Computational Linguistics.