@inproceedings{kim-etal-2026-trinity,
title = "Trinity: Human{--}Agent{--}Environment Alignment for Embodied Task Planning",
author = "Kim, Hyungmin and
Jeon, Hobeom and
Kim, Dohyung and
Jang, Jinhyeok and
Jang, Minsu",
editor = "Lee, Hung-Yi and
Dev, Sunipa and
Li, Jessy and
Elliott, Desmond",
booktitle = "Findings of the {A}ssociation for {C}omputational {L}inguistics: {EMNLP} 2026",
month = oct,
year = "2026",
address = "Budapest, Hungary",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.findings-emnlp.237/",
pages = "4974--5013",
ISBN = "979-8-89176-462-0",
abstract = "LLM-based embodied agents are increasingly evaluated on environment-grounded, long-horizon planning, yet existing benchmarks rarely test whether an agent can condition its plans on the specific user it serves. This gap points to a broader limitation in embodied task-planning systems, which often ground their reasoning only in the physical environment while overlooking the diversity of the people they serve. We frame this problem as Planning with Human Attributes (PHA), in which an agent must integrate explicit user information into long-horizon embodied task planning. We introduce Trinity, a benchmark designed to evaluate PHA capabilities. Within Trinity, a single instruction can yield different, equally valid, or even opposing outcomes depending on a user{'}s personal attributes. Our experiments reveal that standard off-the-shelf in-context learning-based agents commonly fail at the human-grounding component, whether or not they are provided with relevant examples, highlighting a key obstacle for PHA tasks. Beyond benchmarking, we contribute a bottom-up failure-mode analysis that isolates where and why agents break, and we distill it into concrete, model-agnostic design principles for human-agent-environment (HAE) aligned agents. We hope Trinity brings domestic robots a step closer to becoming trusted, socially acceptable partners."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="kim-etal-2026-trinity">
<titleInfo>
<title>Trinity: Human–Agent–Environment Alignment for Embodied Task Planning</title>
</titleInfo>
<name type="personal">
<namePart type="given">Hyungmin</namePart>
<namePart type="family">Kim</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Hobeom</namePart>
<namePart type="family">Jeon</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Dohyung</namePart>
<namePart type="family">Kim</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jinhyeok</namePart>
<namePart type="family">Jang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Minsu</namePart>
<namePart type="family">Jang</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>Findings of the Association for Computational Linguistics: EMNLP 2026</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">979-8-89176-462-0</identifier>
</relatedItem>
<abstract>LLM-based embodied agents are increasingly evaluated on environment-grounded, long-horizon planning, yet existing benchmarks rarely test whether an agent can condition its plans on the specific user it serves. This gap points to a broader limitation in embodied task-planning systems, which often ground their reasoning only in the physical environment while overlooking the diversity of the people they serve. We frame this problem as Planning with Human Attributes (PHA), in which an agent must integrate explicit user information into long-horizon embodied task planning. We introduce Trinity, a benchmark designed to evaluate PHA capabilities. Within Trinity, a single instruction can yield different, equally valid, or even opposing outcomes depending on a user’s personal attributes. Our experiments reveal that standard off-the-shelf in-context learning-based agents commonly fail at the human-grounding component, whether or not they are provided with relevant examples, highlighting a key obstacle for PHA tasks. Beyond benchmarking, we contribute a bottom-up failure-mode analysis that isolates where and why agents break, and we distill it into concrete, model-agnostic design principles for human-agent-environment (HAE) aligned agents. We hope Trinity brings domestic robots a step closer to becoming trusted, socially acceptable partners.</abstract>
<identifier type="citekey">kim-etal-2026-trinity</identifier>
<location>
<url>https://aclanthology.org/2026.findings-emnlp.237/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>4974</start>
<end>5013</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Trinity: Human–Agent–Environment Alignment for Embodied Task Planning
%A Kim, Hyungmin
%A Jeon, Hobeom
%A Kim, Dohyung
%A Jang, Jinhyeok
%A Jang, Minsu
%Y Lee, Hung-Yi
%Y Dev, Sunipa
%Y Li, Jessy
%Y Elliott, Desmond
%S Findings of the Association for Computational Linguistics: EMNLP 2026
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-462-0
%F kim-etal-2026-trinity
%X LLM-based embodied agents are increasingly evaluated on environment-grounded, long-horizon planning, yet existing benchmarks rarely test whether an agent can condition its plans on the specific user it serves. This gap points to a broader limitation in embodied task-planning systems, which often ground their reasoning only in the physical environment while overlooking the diversity of the people they serve. We frame this problem as Planning with Human Attributes (PHA), in which an agent must integrate explicit user information into long-horizon embodied task planning. We introduce Trinity, a benchmark designed to evaluate PHA capabilities. Within Trinity, a single instruction can yield different, equally valid, or even opposing outcomes depending on a user’s personal attributes. Our experiments reveal that standard off-the-shelf in-context learning-based agents commonly fail at the human-grounding component, whether or not they are provided with relevant examples, highlighting a key obstacle for PHA tasks. Beyond benchmarking, we contribute a bottom-up failure-mode analysis that isolates where and why agents break, and we distill it into concrete, model-agnostic design principles for human-agent-environment (HAE) aligned agents. We hope Trinity brings domestic robots a step closer to becoming trusted, socially acceptable partners.
%U https://aclanthology.org/2026.findings-emnlp.237/
%P 4974-5013
Markdown (Informal)
[Trinity: Human–Agent–Environment Alignment for Embodied Task Planning](https://aclanthology.org/2026.findings-emnlp.237/) (Kim et al., Findings 2026)
ACL