@inproceedings{jung-etal-2025-diatool,
title = "{D}ia{T}ool-{DPO}: Multi-Turn Direct Preference Optimization for Tool-Augmented Large Language Models",
author = "Jung, Sunghee and
Lee, Donghun and
Lee, Shinbok and
Seo, Gaeun and
Lee, Daniel and
Ko, Byeongil and
Cho, Junrae and
Kim, Kihyun and
Kim, EungGyun and
Shin, Myeongcheol",
editor = "B{\'e}chet, Fr{\'e}d{\'e}ric and
Lef{\`e}vre, Fabrice and
Asher, Nicholas and
Kim, Seokhwan and
Merlin, Teva",
booktitle = "Proceedings of the 26th Annual Meeting of the Special Interest Group on Discourse and Dialogue",
month = aug,
year = "2025",
address = "Avignon, France",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2025.sigdial-1.32/",
pages = "397--416",
abstract = "Tool-Augmented Larage Language Models (TA-LLMs) have shown promise in real-world applications, but face challenges in handling incomplete queries and out-of-scope requests. While existing approaches rely mainly on Supervised Fine-Tuning with expert trajectories, we propose DiaTool-DPO, a novel method that enhances TA-LLM{'}s dialogue capabilities through Direct Preference Optimization. We model TA-LLM interactions as a Markov Decision Process with 5 distinct dialogue states and categorize user queries into 3 types based on their state transition trajectories. We automatically construct paired trajectory datasets of correct and incorrect dialogue flows and introduce a specialized objective loss for dialogue control. Our comprehensive evaluation demonstrates that DiaTool-DPO approaches GPT-4o{'}s performance (94.8{\%} in information gathering, 91{\%} in tool call rejection) with substantial improvements over baseline (44{\%} and 9.6{\%} respectively) while maintaining core functionality. Our approach opens new possibilities for developing TA-LLMs that can handle diverse real-world scenarios without requiring additional expert demonstrations or human labeling."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="jung-etal-2025-diatool">
<titleInfo>
<title>DiaTool-DPO: Multi-Turn Direct Preference Optimization for Tool-Augmented Large Language Models</title>
</titleInfo>
<name type="personal">
<namePart type="given">Sunghee</namePart>
<namePart type="family">Jung</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Donghun</namePart>
<namePart type="family">Lee</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Shinbok</namePart>
<namePart type="family">Lee</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Gaeun</namePart>
<namePart type="family">Seo</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Daniel</namePart>
<namePart type="family">Lee</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Byeongil</namePart>
<namePart type="family">Ko</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Junrae</namePart>
<namePart type="family">Cho</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Kihyun</namePart>
<namePart type="family">Kim</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">EungGyun</namePart>
<namePart type="family">Kim</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Myeongcheol</namePart>
<namePart type="family">Shin</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2025-08</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 26th Annual Meeting of the Special Interest Group on Discourse and Dialogue</title>
</titleInfo>
<name type="personal">
<namePart type="given">Frédéric</namePart>
<namePart type="family">Béchet</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Fabrice</namePart>
<namePart type="family">Lefèvre</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Nicholas</namePart>
<namePart type="family">Asher</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Seokhwan</namePart>
<namePart type="family">Kim</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Teva</namePart>
<namePart type="family">Merlin</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Avignon, France</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>Tool-Augmented Larage Language Models (TA-LLMs) have shown promise in real-world applications, but face challenges in handling incomplete queries and out-of-scope requests. While existing approaches rely mainly on Supervised Fine-Tuning with expert trajectories, we propose DiaTool-DPO, a novel method that enhances TA-LLM’s dialogue capabilities through Direct Preference Optimization. We model TA-LLM interactions as a Markov Decision Process with 5 distinct dialogue states and categorize user queries into 3 types based on their state transition trajectories. We automatically construct paired trajectory datasets of correct and incorrect dialogue flows and introduce a specialized objective loss for dialogue control. Our comprehensive evaluation demonstrates that DiaTool-DPO approaches GPT-4o’s performance (94.8% in information gathering, 91% in tool call rejection) with substantial improvements over baseline (44% and 9.6% respectively) while maintaining core functionality. Our approach opens new possibilities for developing TA-LLMs that can handle diverse real-world scenarios without requiring additional expert demonstrations or human labeling.</abstract>
<identifier type="citekey">jung-etal-2025-diatool</identifier>
<location>
<url>https://aclanthology.org/2025.sigdial-1.32/</url>
</location>
<part>
<date>2025-08</date>
<extent unit="page">
<start>397</start>
<end>416</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T DiaTool-DPO: Multi-Turn Direct Preference Optimization for Tool-Augmented Large Language Models
%A Jung, Sunghee
%A Lee, Donghun
%A Lee, Shinbok
%A Seo, Gaeun
%A Lee, Daniel
%A Ko, Byeongil
%A Cho, Junrae
%A Kim, Kihyun
%A Kim, EungGyun
%A Shin, Myeongcheol
%Y Béchet, Frédéric
%Y Lefèvre, Fabrice
%Y Asher, Nicholas
%Y Kim, Seokhwan
%Y Merlin, Teva
%S Proceedings of the 26th Annual Meeting of the Special Interest Group on Discourse and Dialogue
%D 2025
%8 August
%I Association for Computational Linguistics
%C Avignon, France
%F jung-etal-2025-diatool
%X Tool-Augmented Larage Language Models (TA-LLMs) have shown promise in real-world applications, but face challenges in handling incomplete queries and out-of-scope requests. While existing approaches rely mainly on Supervised Fine-Tuning with expert trajectories, we propose DiaTool-DPO, a novel method that enhances TA-LLM’s dialogue capabilities through Direct Preference Optimization. We model TA-LLM interactions as a Markov Decision Process with 5 distinct dialogue states and categorize user queries into 3 types based on their state transition trajectories. We automatically construct paired trajectory datasets of correct and incorrect dialogue flows and introduce a specialized objective loss for dialogue control. Our comprehensive evaluation demonstrates that DiaTool-DPO approaches GPT-4o’s performance (94.8% in information gathering, 91% in tool call rejection) with substantial improvements over baseline (44% and 9.6% respectively) while maintaining core functionality. Our approach opens new possibilities for developing TA-LLMs that can handle diverse real-world scenarios without requiring additional expert demonstrations or human labeling.
%U https://aclanthology.org/2025.sigdial-1.32/
%P 397-416
Markdown (Informal)
[DiaTool-DPO: Multi-Turn Direct Preference Optimization for Tool-Augmented Large Language Models](https://aclanthology.org/2025.sigdial-1.32/) (Jung et al., SIGDIAL 2025)
ACL
- Sunghee Jung, Donghun Lee, Shinbok Lee, Gaeun Seo, Daniel Lee, Byeongil Ko, Junrae Cho, Kihyun Kim, EungGyun Kim, and Myeongcheol Shin. 2025. DiaTool-DPO: Multi-Turn Direct Preference Optimization for Tool-Augmented Large Language Models. In Proceedings of the 26th Annual Meeting of the Special Interest Group on Discourse and Dialogue, pages 397–416, Avignon, France. Association for Computational Linguistics.