@inproceedings{chen-xuan-qi-etal-2026-trac,
title = "{TRAC}: Token-level Reward Assignment for Coherent Abstractive Summarization",
author = "陈宣齐 and
容梓莹 and
Liao, Xinfeng and
Wang, Lianxi and
Gao, Ying and
Jiang, Shengyi",
editor = "Liakata, Maria and
Moreira, Viviane P. and
Zhang, Jiajun and
Jurgens, David",
booktitle = "Findings of the {A}ssociation for {C}omputational {L}inguistics: {ACL} 2026",
month = jul,
year = "2026",
address = "San Diego, California, United States",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.findings-acl.1873/",
doi = "10.18653/v1/2026.findings-acl.1873",
pages = "37563--37578",
ISBN = "979-8-89176-395-1",
abstract = "Large Language Models (LLMs) have achieved remarkable success in text summarization, particularly through the integration of reinforcement learning. However, maintaining logical coherence and contextual consistency remains a pervasive challenge in long-form generation, often hindering the production of high-quality, unified summaries. To address these persistent issues, we propose TRAC, a framework that introduces a token-level reward function by integrating relative sentence gain, inter-sentence attention, and a Gaussian length penalty. By training a Process Reward Model (PRM) to provide fine-grained, step-wise supervision, TRAC ensures superior structural integrity and fluency during the generation process. Experimental results demonstrate that TRAC outperforms the sequence-level baseline by 11.05{\%} in Fluency and 10.61{\%} in Relevance. Furthermore, it achieves significant gains over competitive baselines such as FIGA and TLCR, underscoring its effectiveness and generalizability in high-quality NLP summarization."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="chen-xuan-qi-etal-2026-trac">
<titleInfo>
<title>TRAC: Token-level Reward Assignment for Coherent Abstractive Summarization</title>
</titleInfo>
<name>
<namePart>陈宣齐</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name>
<namePart>容梓莹</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Xinfeng</namePart>
<namePart type="family">Liao</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Lianxi</namePart>
<namePart type="family">Wang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ying</namePart>
<namePart type="family">Gao</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Shengyi</namePart>
<namePart type="family">Jiang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-07</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Findings of the Association for Computational Linguistics: ACL 2026</title>
</titleInfo>
<name type="personal">
<namePart type="given">Maria</namePart>
<namePart type="family">Liakata</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Viviane</namePart>
<namePart type="given">P</namePart>
<namePart type="family">Moreira</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jiajun</namePart>
<namePart type="family">Zhang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">David</namePart>
<namePart type="family">Jurgens</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">San Diego, California, United States</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-395-1</identifier>
</relatedItem>
<abstract>Large Language Models (LLMs) have achieved remarkable success in text summarization, particularly through the integration of reinforcement learning. However, maintaining logical coherence and contextual consistency remains a pervasive challenge in long-form generation, often hindering the production of high-quality, unified summaries. To address these persistent issues, we propose TRAC, a framework that introduces a token-level reward function by integrating relative sentence gain, inter-sentence attention, and a Gaussian length penalty. By training a Process Reward Model (PRM) to provide fine-grained, step-wise supervision, TRAC ensures superior structural integrity and fluency during the generation process. Experimental results demonstrate that TRAC outperforms the sequence-level baseline by 11.05% in Fluency and 10.61% in Relevance. Furthermore, it achieves significant gains over competitive baselines such as FIGA and TLCR, underscoring its effectiveness and generalizability in high-quality NLP summarization.</abstract>
<identifier type="citekey">chen-xuan-qi-etal-2026-trac</identifier>
<identifier type="doi">10.18653/v1/2026.findings-acl.1873</identifier>
<location>
<url>https://aclanthology.org/2026.findings-acl.1873/</url>
</location>
<part>
<date>2026-07</date>
<extent unit="page">
<start>37563</start>
<end>37578</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T TRAC: Token-level Reward Assignment for Coherent Abstractive Summarization
%A Liao, Xinfeng
%A Wang, Lianxi
%A Gao, Ying
%A Jiang, Shengyi
%Y Liakata, Maria
%Y Moreira, Viviane P.
%Y Zhang, Jiajun
%Y Jurgens, David
%A 陈宣齐
%A 容梓莹
%S Findings of the Association for Computational Linguistics: ACL 2026
%D 2026
%8 July
%I Association for Computational Linguistics
%C San Diego, California, United States
%@ 979-8-89176-395-1
%F chen-xuan-qi-etal-2026-trac
%X Large Language Models (LLMs) have achieved remarkable success in text summarization, particularly through the integration of reinforcement learning. However, maintaining logical coherence and contextual consistency remains a pervasive challenge in long-form generation, often hindering the production of high-quality, unified summaries. To address these persistent issues, we propose TRAC, a framework that introduces a token-level reward function by integrating relative sentence gain, inter-sentence attention, and a Gaussian length penalty. By training a Process Reward Model (PRM) to provide fine-grained, step-wise supervision, TRAC ensures superior structural integrity and fluency during the generation process. Experimental results demonstrate that TRAC outperforms the sequence-level baseline by 11.05% in Fluency and 10.61% in Relevance. Furthermore, it achieves significant gains over competitive baselines such as FIGA and TLCR, underscoring its effectiveness and generalizability in high-quality NLP summarization.
%R 10.18653/v1/2026.findings-acl.1873
%U https://aclanthology.org/2026.findings-acl.1873/
%U https://doi.org/10.18653/v1/2026.findings-acl.1873
%P 37563-37578
Markdown (Informal)
[TRAC: Token-level Reward Assignment for Coherent Abstractive Summarization](https://aclanthology.org/2026.findings-acl.1873/) (陈宣齐 et al., Findings 2026)
ACL