@inproceedings{yue-etal-2026-navigating,
title = "Navigating Intention Reasoning Trees to Guide Agentic Preference Optimization",
author = "Yue, Ziben and
Yang, Hongyu and
Li, Weixin",
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.1740/",
doi = "10.18653/2026.findings-emnlp.1740",
pages = "37209--37228",
ISBN = "979-8-89176-462-0",
abstract = "Understanding human intentions in videos requires deep cognitive reasoning, yet Multimodal Large Language Models (MLLMs) frequently suffer from factual hallucinations and counterfactual blindness. To address these limitations, we introduce ASTRO, a novel framework that navigates Intention Reansoning Tree to guide preference optimization. Our approach employs a collaborative multi-agent system integrated with Monte Carlo Tree Search (MCTS) to systematically explore complex reasoning spaces based on visual evidence. By navigating this tree, ASTRO dynamically extracts distinct cognitive trajectories{---}correct, error, and ambiguous paths{---}to construct high-quality preference pairs. Finally, we apply Direct Preference Optimization (DPO) using these contrastive paths, explicitly enhancing the model{'}s factual grounding and sensitivity to causal boundaries. Extensive experiments across complex video intention benchmarks demonstrate that ASTRO achieves superior performance while exhibiting remarkable data efficiency, significantly outperforming fully supervised baselines using only a fraction of the original training data."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="yue-etal-2026-navigating">
<titleInfo>
<title>Navigating Intention Reasoning Trees to Guide Agentic Preference Optimization</title>
</titleInfo>
<name type="personal">
<namePart type="given">Ziben</namePart>
<namePart type="family">Yue</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Hongyu</namePart>
<namePart type="family">Yang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Weixin</namePart>
<namePart type="family">Li</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>Understanding human intentions in videos requires deep cognitive reasoning, yet Multimodal Large Language Models (MLLMs) frequently suffer from factual hallucinations and counterfactual blindness. To address these limitations, we introduce ASTRO, a novel framework that navigates Intention Reansoning Tree to guide preference optimization. Our approach employs a collaborative multi-agent system integrated with Monte Carlo Tree Search (MCTS) to systematically explore complex reasoning spaces based on visual evidence. By navigating this tree, ASTRO dynamically extracts distinct cognitive trajectories—correct, error, and ambiguous paths—to construct high-quality preference pairs. Finally, we apply Direct Preference Optimization (DPO) using these contrastive paths, explicitly enhancing the model’s factual grounding and sensitivity to causal boundaries. Extensive experiments across complex video intention benchmarks demonstrate that ASTRO achieves superior performance while exhibiting remarkable data efficiency, significantly outperforming fully supervised baselines using only a fraction of the original training data.</abstract>
<identifier type="citekey">yue-etal-2026-navigating</identifier>
<identifier type="doi">10.18653/2026.findings-emnlp.1740</identifier>
<location>
<url>https://aclanthology.org/2026.findings-emnlp.1740/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>37209</start>
<end>37228</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Navigating Intention Reasoning Trees to Guide Agentic Preference Optimization
%A Yue, Ziben
%A Yang, Hongyu
%A Li, Weixin
%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 yue-etal-2026-navigating
%X Understanding human intentions in videos requires deep cognitive reasoning, yet Multimodal Large Language Models (MLLMs) frequently suffer from factual hallucinations and counterfactual blindness. To address these limitations, we introduce ASTRO, a novel framework that navigates Intention Reansoning Tree to guide preference optimization. Our approach employs a collaborative multi-agent system integrated with Monte Carlo Tree Search (MCTS) to systematically explore complex reasoning spaces based on visual evidence. By navigating this tree, ASTRO dynamically extracts distinct cognitive trajectories—correct, error, and ambiguous paths—to construct high-quality preference pairs. Finally, we apply Direct Preference Optimization (DPO) using these contrastive paths, explicitly enhancing the model’s factual grounding and sensitivity to causal boundaries. Extensive experiments across complex video intention benchmarks demonstrate that ASTRO achieves superior performance while exhibiting remarkable data efficiency, significantly outperforming fully supervised baselines using only a fraction of the original training data.
%R 10.18653/2026.findings-emnlp.1740
%U https://aclanthology.org/2026.findings-emnlp.1740/
%U https://doi.org/10.18653/2026.findings-emnlp.1740
%P 37209-37228
Markdown (Informal)
[Navigating Intention Reasoning Trees to Guide Agentic Preference Optimization](https://aclanthology.org/2026.findings-emnlp.1740/) (Yue et al., Findings 2026)
ACL