@inproceedings{barta-etal-2025-enhancing,
title = "Enhancing {AMR} Parsing with Group Relative Policy Optimization",
author = "Barta, Botond and
Hamerlik, Endre and
Nyist, Mil{\'a}n and
Ito, Masato and
Acs, Judit",
editor = "Fei, Hao and
Tu, Kewei and
Zhang, Yuhui and
Hu, Xiang and
Han, Wenjuan and
Jia, Zixia and
Zheng, Zilong and
Cao, Yixin and
Zhang, Meishan and
Lu, Wei and
Siddharth, N. and
{\O}vrelid, Lilja and
Xue, Nianwen and
Zhang, Yue",
booktitle = "Proceedings of the 1st Joint Workshop on Large Language Models and Structure Modeling (XLLM 2025)",
month = aug,
year = "2025",
address = "Vienna, Austria",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2025.xllm-1.11/",
doi = "10.18653/v1/2025.xllm-1.11",
pages = "99--105",
ISBN = "979-8-89176-286-2",
abstract = "We investigate the capabilities of the openly available Llama 3.2 1B language model for Abstract Meaning Representation (AMR) parsing through supervised fine-tuning, further enhanced by reinforcement learning via Group Relative Policy Optimization (GRPO). Existing supervised methods for AMR parsing face limitations due to static loss functions and challenges in capturing complex semantic phenomena. To address this, our GRPO-based approach explicitly optimizes fine-grained semantic rewards, including Smatch scores, frame-argument correctness, and structural validity of logical operations. Experimental results show that supervised fine-tuning alone establishes Llama as a capable English AMR parser, and subsequent GRPO fine-tuning further improves its performance. Our final model achieves higher Smatch scores, consistently respects critical low-level semantic constraints, and outperforms existing parsers on high-level semantic evaluation metrics across diverse linguistic phenomena."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="barta-etal-2025-enhancing">
<titleInfo>
<title>Enhancing AMR Parsing with Group Relative Policy Optimization</title>
</titleInfo>
<name type="personal">
<namePart type="given">Botond</namePart>
<namePart type="family">Barta</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Endre</namePart>
<namePart type="family">Hamerlik</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Milán</namePart>
<namePart type="family">Nyist</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Masato</namePart>
<namePart type="family">Ito</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Judit</namePart>
<namePart type="family">Acs</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 1st Joint Workshop on Large Language Models and Structure Modeling (XLLM 2025)</title>
</titleInfo>
<name type="personal">
<namePart type="given">Hao</namePart>
<namePart type="family">Fei</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Kewei</namePart>
<namePart type="family">Tu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yuhui</namePart>
<namePart type="family">Zhang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Xiang</namePart>
<namePart type="family">Hu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Wenjuan</namePart>
<namePart type="family">Han</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Zixia</namePart>
<namePart type="family">Jia</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Zilong</namePart>
<namePart type="family">Zheng</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yixin</namePart>
<namePart type="family">Cao</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Meishan</namePart>
<namePart type="family">Zhang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Wei</namePart>
<namePart type="family">Lu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">N</namePart>
<namePart type="family">Siddharth</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Lilja</namePart>
<namePart type="family">Øvrelid</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Nianwen</namePart>
<namePart type="family">Xue</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yue</namePart>
<namePart type="family">Zhang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Vienna, Austria</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-286-2</identifier>
</relatedItem>
<abstract>We investigate the capabilities of the openly available Llama 3.2 1B language model for Abstract Meaning Representation (AMR) parsing through supervised fine-tuning, further enhanced by reinforcement learning via Group Relative Policy Optimization (GRPO). Existing supervised methods for AMR parsing face limitations due to static loss functions and challenges in capturing complex semantic phenomena. To address this, our GRPO-based approach explicitly optimizes fine-grained semantic rewards, including Smatch scores, frame-argument correctness, and structural validity of logical operations. Experimental results show that supervised fine-tuning alone establishes Llama as a capable English AMR parser, and subsequent GRPO fine-tuning further improves its performance. Our final model achieves higher Smatch scores, consistently respects critical low-level semantic constraints, and outperforms existing parsers on high-level semantic evaluation metrics across diverse linguistic phenomena.</abstract>
<identifier type="citekey">barta-etal-2025-enhancing</identifier>
<identifier type="doi">10.18653/v1/2025.xllm-1.11</identifier>
<location>
<url>https://aclanthology.org/2025.xllm-1.11/</url>
</location>
<part>
<date>2025-08</date>
<extent unit="page">
<start>99</start>
<end>105</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Enhancing AMR Parsing with Group Relative Policy Optimization
%A Barta, Botond
%A Hamerlik, Endre
%A Nyist, Milán
%A Ito, Masato
%A Acs, Judit
%Y Fei, Hao
%Y Tu, Kewei
%Y Zhang, Yuhui
%Y Hu, Xiang
%Y Han, Wenjuan
%Y Jia, Zixia
%Y Zheng, Zilong
%Y Cao, Yixin
%Y Zhang, Meishan
%Y Lu, Wei
%Y Siddharth, N.
%Y Øvrelid, Lilja
%Y Xue, Nianwen
%Y Zhang, Yue
%S Proceedings of the 1st Joint Workshop on Large Language Models and Structure Modeling (XLLM 2025)
%D 2025
%8 August
%I Association for Computational Linguistics
%C Vienna, Austria
%@ 979-8-89176-286-2
%F barta-etal-2025-enhancing
%X We investigate the capabilities of the openly available Llama 3.2 1B language model for Abstract Meaning Representation (AMR) parsing through supervised fine-tuning, further enhanced by reinforcement learning via Group Relative Policy Optimization (GRPO). Existing supervised methods for AMR parsing face limitations due to static loss functions and challenges in capturing complex semantic phenomena. To address this, our GRPO-based approach explicitly optimizes fine-grained semantic rewards, including Smatch scores, frame-argument correctness, and structural validity of logical operations. Experimental results show that supervised fine-tuning alone establishes Llama as a capable English AMR parser, and subsequent GRPO fine-tuning further improves its performance. Our final model achieves higher Smatch scores, consistently respects critical low-level semantic constraints, and outperforms existing parsers on high-level semantic evaluation metrics across diverse linguistic phenomena.
%R 10.18653/v1/2025.xllm-1.11
%U https://aclanthology.org/2025.xllm-1.11/
%U https://doi.org/10.18653/v1/2025.xllm-1.11
%P 99-105
Markdown (Informal)
[Enhancing AMR Parsing with Group Relative Policy Optimization](https://aclanthology.org/2025.xllm-1.11/) (Barta et al., XLLM 2025)
ACL
- Botond Barta, Endre Hamerlik, Milán Nyist, Masato Ito, and Judit Acs. 2025. Enhancing AMR Parsing with Group Relative Policy Optimization. In Proceedings of the 1st Joint Workshop on Large Language Models and Structure Modeling (XLLM 2025), pages 99–105, Vienna, Austria. Association for Computational Linguistics.