@inproceedings{akarajaradwong-etal-2025-aligning,
title = "Aligning {LLM}s for {T}hai Legal Question Answering with Efficient Semantic-Similarity Rewards",
author = "Akarajaradwong, Pawitsapak and
Chaksangchaichot, Chompakorn and
Pothavorn, Pirat and
Chuangsuwanich, Ekapol and
Rutherford, Attapol and
Nutanong, Sarana",
editor = "Aletras, Nikolaos and
Chalkidis, Ilias and
Barrett, Leslie and
Goanț{\u{a}}, C{\u{a}}t{\u{a}}lina and
Preoțiuc-Pietro, Daniel and
Spanakis, Gerasimos",
booktitle = "Proceedings of the Natural Legal Language Processing Workshop 2025",
month = nov,
year = "2025",
address = "Suzhou, China",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2025.nllp-1.21/",
pages = "304--316",
ISBN = "979-8-89176-338-8",
abstract = "The Retrieval-Augmented Generation (RAG) systems' performance on Thai legal question answering is still limited, especially for questions requiring extensive, complex legal reasoning. To address these limitations, we introduce a resource-efficient approach that aligns Large Language Models (LLMs) for improved citation accuracy and response quality using Group-Relative Policy Optimization (GRPO). Our proposed method leverages BGE-M3 embeddings as a cost-efficient semantic-similarity reward, significantly reducing computational expenses up to 2.5x compared to an LLM-based reward model. Experiments on the NitiBench benchmark demonstrate substantial improvements: GRPO achieves up to 90{\%} citation-F1 gains relative to the base model and a 31{\%} increase in joint quality metrics over instruction tuning. Crucially, our approach provides a practical and effective solution for enhancing legal LLMs in resource-constrained environments."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="akarajaradwong-etal-2025-aligning">
<titleInfo>
<title>Aligning LLMs for Thai Legal Question Answering with Efficient Semantic-Similarity Rewards</title>
</titleInfo>
<name type="personal">
<namePart type="given">Pawitsapak</namePart>
<namePart type="family">Akarajaradwong</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Chompakorn</namePart>
<namePart type="family">Chaksangchaichot</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Pirat</namePart>
<namePart type="family">Pothavorn</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ekapol</namePart>
<namePart type="family">Chuangsuwanich</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Attapol</namePart>
<namePart type="family">Rutherford</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Sarana</namePart>
<namePart type="family">Nutanong</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2025-11</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the Natural Legal Language Processing Workshop 2025</title>
</titleInfo>
<name type="personal">
<namePart type="given">Nikolaos</namePart>
<namePart type="family">Aletras</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ilias</namePart>
<namePart type="family">Chalkidis</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Leslie</namePart>
<namePart type="family">Barrett</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Cătălina</namePart>
<namePart type="family">Goanță</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Daniel</namePart>
<namePart type="family">Preoțiuc-Pietro</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Gerasimos</namePart>
<namePart type="family">Spanakis</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Suzhou, China</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-338-8</identifier>
</relatedItem>
<abstract>The Retrieval-Augmented Generation (RAG) systems’ performance on Thai legal question answering is still limited, especially for questions requiring extensive, complex legal reasoning. To address these limitations, we introduce a resource-efficient approach that aligns Large Language Models (LLMs) for improved citation accuracy and response quality using Group-Relative Policy Optimization (GRPO). Our proposed method leverages BGE-M3 embeddings as a cost-efficient semantic-similarity reward, significantly reducing computational expenses up to 2.5x compared to an LLM-based reward model. Experiments on the NitiBench benchmark demonstrate substantial improvements: GRPO achieves up to 90% citation-F1 gains relative to the base model and a 31% increase in joint quality metrics over instruction tuning. Crucially, our approach provides a practical and effective solution for enhancing legal LLMs in resource-constrained environments.</abstract>
<identifier type="citekey">akarajaradwong-etal-2025-aligning</identifier>
<location>
<url>https://aclanthology.org/2025.nllp-1.21/</url>
</location>
<part>
<date>2025-11</date>
<extent unit="page">
<start>304</start>
<end>316</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Aligning LLMs for Thai Legal Question Answering with Efficient Semantic-Similarity Rewards
%A Akarajaradwong, Pawitsapak
%A Chaksangchaichot, Chompakorn
%A Pothavorn, Pirat
%A Chuangsuwanich, Ekapol
%A Rutherford, Attapol
%A Nutanong, Sarana
%Y Aletras, Nikolaos
%Y Chalkidis, Ilias
%Y Barrett, Leslie
%Y Goanță, Cătălina
%Y Preoțiuc-Pietro, Daniel
%Y Spanakis, Gerasimos
%S Proceedings of the Natural Legal Language Processing Workshop 2025
%D 2025
%8 November
%I Association for Computational Linguistics
%C Suzhou, China
%@ 979-8-89176-338-8
%F akarajaradwong-etal-2025-aligning
%X The Retrieval-Augmented Generation (RAG) systems’ performance on Thai legal question answering is still limited, especially for questions requiring extensive, complex legal reasoning. To address these limitations, we introduce a resource-efficient approach that aligns Large Language Models (LLMs) for improved citation accuracy and response quality using Group-Relative Policy Optimization (GRPO). Our proposed method leverages BGE-M3 embeddings as a cost-efficient semantic-similarity reward, significantly reducing computational expenses up to 2.5x compared to an LLM-based reward model. Experiments on the NitiBench benchmark demonstrate substantial improvements: GRPO achieves up to 90% citation-F1 gains relative to the base model and a 31% increase in joint quality metrics over instruction tuning. Crucially, our approach provides a practical and effective solution for enhancing legal LLMs in resource-constrained environments.
%U https://aclanthology.org/2025.nllp-1.21/
%P 304-316
Markdown (Informal)
[Aligning LLMs for Thai Legal Question Answering with Efficient Semantic-Similarity Rewards](https://aclanthology.org/2025.nllp-1.21/) (Akarajaradwong et al., NLLP 2025)
ACL