@inproceedings{wei-etal-2025-cnnsum,
title = "{CNNS}um: Exploring Long-Context Summarization with Large Language Models in {C}hinese Novels",
author = "Wei, Lingxiao and
Yan, He and
Xiangju, Lu and
Zhu, Junmin and
Wang, Jun and
Zhang, Wei",
editor = "Che, Wanxiang and
Nabende, Joyce and
Shutova, Ekaterina and
Pilehvar, Mohammad Taher",
booktitle = "Findings of the Association for Computational Linguistics: ACL 2025",
month = jul,
year = "2025",
address = "Vienna, Austria",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2025.findings-acl.421/",
doi = "10.18653/v1/2025.findings-acl.421",
pages = "8034--8062",
ISBN = "979-8-89176-256-5",
abstract = "Large language models (LLMs) have been well-researched in various long-context tasks. However, the scarcity of long-context summarization datasets hinders progress in this area. To address this, we introduce CNNSum, a multi-scale long-context summarization benchmark based on Chinese novels, featuring human-driven annotations across four subsets totaling 695 samples, with lengths ranging from 16k to 128k. We benchmark numerous LLMs and conduct detailed human assessments to summarize abnormal output types. Furthermore, we extensively explore how to improve long-context summarization. In our study: (1) Advanced LLMs may generate much subjective commentary, leading to vague summaries. (2) Currently, long-context summarization mainly relies on memory ability. The advantages of Large LLMs are hard to utilize, thus small LLMs are more cost-effective. (3) Different prompt types paired with various version models may cause large performance gaps. In further fine-tuning, these can be mitigated, and the Base version models perform better. (4) LLMs with RoPE-base scaled exhibit strong extrapolation potential; using short-context data can significantly improve long-context summarization performance. However, further applying other interpolation methods requires careful selection. (5) CNNSum provides more reliable evaluation results than other benchmarks. We release CNNSum to advance future research."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="wei-etal-2025-cnnsum">
<titleInfo>
<title>CNNSum: Exploring Long-Context Summarization with Large Language Models in Chinese Novels</title>
</titleInfo>
<name type="personal">
<namePart type="given">Lingxiao</namePart>
<namePart type="family">Wei</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">He</namePart>
<namePart type="family">Yan</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Lu</namePart>
<namePart type="family">Xiangju</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Junmin</namePart>
<namePart type="family">Zhu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jun</namePart>
<namePart type="family">Wang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Wei</namePart>
<namePart type="family">Zhang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2025-07</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Findings of the Association for Computational Linguistics: ACL 2025</title>
</titleInfo>
<name type="personal">
<namePart type="given">Wanxiang</namePart>
<namePart type="family">Che</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Joyce</namePart>
<namePart type="family">Nabende</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ekaterina</namePart>
<namePart type="family">Shutova</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Mohammad</namePart>
<namePart type="given">Taher</namePart>
<namePart type="family">Pilehvar</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">979-8-89176-256-5</identifier>
</relatedItem>
<abstract>Large language models (LLMs) have been well-researched in various long-context tasks. However, the scarcity of long-context summarization datasets hinders progress in this area. To address this, we introduce CNNSum, a multi-scale long-context summarization benchmark based on Chinese novels, featuring human-driven annotations across four subsets totaling 695 samples, with lengths ranging from 16k to 128k. We benchmark numerous LLMs and conduct detailed human assessments to summarize abnormal output types. Furthermore, we extensively explore how to improve long-context summarization. In our study: (1) Advanced LLMs may generate much subjective commentary, leading to vague summaries. (2) Currently, long-context summarization mainly relies on memory ability. The advantages of Large LLMs are hard to utilize, thus small LLMs are more cost-effective. (3) Different prompt types paired with various version models may cause large performance gaps. In further fine-tuning, these can be mitigated, and the Base version models perform better. (4) LLMs with RoPE-base scaled exhibit strong extrapolation potential; using short-context data can significantly improve long-context summarization performance. However, further applying other interpolation methods requires careful selection. (5) CNNSum provides more reliable evaluation results than other benchmarks. We release CNNSum to advance future research.</abstract>
<identifier type="citekey">wei-etal-2025-cnnsum</identifier>
<identifier type="doi">10.18653/v1/2025.findings-acl.421</identifier>
<location>
<url>https://aclanthology.org/2025.findings-acl.421/</url>
</location>
<part>
<date>2025-07</date>
<extent unit="page">
<start>8034</start>
<end>8062</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T CNNSum: Exploring Long-Context Summarization with Large Language Models in Chinese Novels
%A Wei, Lingxiao
%A Yan, He
%A Xiangju, Lu
%A Zhu, Junmin
%A Wang, Jun
%A Zhang, Wei
%Y Che, Wanxiang
%Y Nabende, Joyce
%Y Shutova, Ekaterina
%Y Pilehvar, Mohammad Taher
%S Findings of the Association for Computational Linguistics: ACL 2025
%D 2025
%8 July
%I Association for Computational Linguistics
%C Vienna, Austria
%@ 979-8-89176-256-5
%F wei-etal-2025-cnnsum
%X Large language models (LLMs) have been well-researched in various long-context tasks. However, the scarcity of long-context summarization datasets hinders progress in this area. To address this, we introduce CNNSum, a multi-scale long-context summarization benchmark based on Chinese novels, featuring human-driven annotations across four subsets totaling 695 samples, with lengths ranging from 16k to 128k. We benchmark numerous LLMs and conduct detailed human assessments to summarize abnormal output types. Furthermore, we extensively explore how to improve long-context summarization. In our study: (1) Advanced LLMs may generate much subjective commentary, leading to vague summaries. (2) Currently, long-context summarization mainly relies on memory ability. The advantages of Large LLMs are hard to utilize, thus small LLMs are more cost-effective. (3) Different prompt types paired with various version models may cause large performance gaps. In further fine-tuning, these can be mitigated, and the Base version models perform better. (4) LLMs with RoPE-base scaled exhibit strong extrapolation potential; using short-context data can significantly improve long-context summarization performance. However, further applying other interpolation methods requires careful selection. (5) CNNSum provides more reliable evaluation results than other benchmarks. We release CNNSum to advance future research.
%R 10.18653/v1/2025.findings-acl.421
%U https://aclanthology.org/2025.findings-acl.421/
%U https://doi.org/10.18653/v1/2025.findings-acl.421
%P 8034-8062
Markdown (Informal)
[CNNSum: Exploring Long-Context Summarization with Large Language Models in Chinese Novels](https://aclanthology.org/2025.findings-acl.421/) (Wei et al., Findings 2025)
ACL