@inproceedings{dat-etal-2025-discrete,
title = "Discrete Diffusion Language Model for Efficient Text Summarization",
author = "Dat, Do Huu and
Do, Duc Anh and
Luu, Anh Tuan and
Buntine, Wray",
editor = "Chiruzzo, Luis and
Ritter, Alan and
Wang, Lu",
booktitle = "Findings of the Association for Computational Linguistics: NAACL 2025",
month = apr,
year = "2025",
address = "Albuquerque, New Mexico",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2025.findings-naacl.352/",
doi = "10.18653/v1/2025.findings-naacl.352",
pages = "6278--6290",
ISBN = "979-8-89176-195-7",
abstract = "While diffusion models excel at conditionally generating high-quality images, prior works in discrete diffusion models were not evaluated on conditional long-text generation. This work addresses the limitations of prior discrete diffusion models for conditional long-text generation, particularly in the long abstractive summarization task. Despite faster decoding speeds compared to autoregressive methods, previous discrete diffusion models failed on the abstractive summarization task due to the incompatibility between the backbone architectures and the random noising process. To overcome these challenges, we introduce a novel semantic-aware noising process that enables Transformer backbones to handle long sequences effectively. Additionally, we propose CrossMamba, an adaptation of the Mamba model to the encoder-decoder paradigm, which integrates seamlessly with the random absorbing noising process. Our approaches outperform existing discrete diffusion models on three benchmark summarization datasets: Gigaword, CNN/DailyMail, and Arxiv, while also achieving much faster inference speed compared to autoregressive models."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="dat-etal-2025-discrete">
<titleInfo>
<title>Discrete Diffusion Language Model for Efficient Text Summarization</title>
</titleInfo>
<name type="personal">
<namePart type="given">Do</namePart>
<namePart type="given">Huu</namePart>
<namePart type="family">Dat</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Duc</namePart>
<namePart type="given">Anh</namePart>
<namePart type="family">Do</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Anh</namePart>
<namePart type="given">Tuan</namePart>
<namePart type="family">Luu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Wray</namePart>
<namePart type="family">Buntine</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2025-04</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Findings of the Association for Computational Linguistics: NAACL 2025</title>
</titleInfo>
<name type="personal">
<namePart type="given">Luis</namePart>
<namePart type="family">Chiruzzo</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Alan</namePart>
<namePart type="family">Ritter</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Lu</namePart>
<namePart type="family">Wang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Albuquerque, New Mexico</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-195-7</identifier>
</relatedItem>
<abstract>While diffusion models excel at conditionally generating high-quality images, prior works in discrete diffusion models were not evaluated on conditional long-text generation. This work addresses the limitations of prior discrete diffusion models for conditional long-text generation, particularly in the long abstractive summarization task. Despite faster decoding speeds compared to autoregressive methods, previous discrete diffusion models failed on the abstractive summarization task due to the incompatibility between the backbone architectures and the random noising process. To overcome these challenges, we introduce a novel semantic-aware noising process that enables Transformer backbones to handle long sequences effectively. Additionally, we propose CrossMamba, an adaptation of the Mamba model to the encoder-decoder paradigm, which integrates seamlessly with the random absorbing noising process. Our approaches outperform existing discrete diffusion models on three benchmark summarization datasets: Gigaword, CNN/DailyMail, and Arxiv, while also achieving much faster inference speed compared to autoregressive models.</abstract>
<identifier type="citekey">dat-etal-2025-discrete</identifier>
<identifier type="doi">10.18653/v1/2025.findings-naacl.352</identifier>
<location>
<url>https://aclanthology.org/2025.findings-naacl.352/</url>
</location>
<part>
<date>2025-04</date>
<extent unit="page">
<start>6278</start>
<end>6290</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Discrete Diffusion Language Model for Efficient Text Summarization
%A Dat, Do Huu
%A Do, Duc Anh
%A Luu, Anh Tuan
%A Buntine, Wray
%Y Chiruzzo, Luis
%Y Ritter, Alan
%Y Wang, Lu
%S Findings of the Association for Computational Linguistics: NAACL 2025
%D 2025
%8 April
%I Association for Computational Linguistics
%C Albuquerque, New Mexico
%@ 979-8-89176-195-7
%F dat-etal-2025-discrete
%X While diffusion models excel at conditionally generating high-quality images, prior works in discrete diffusion models were not evaluated on conditional long-text generation. This work addresses the limitations of prior discrete diffusion models for conditional long-text generation, particularly in the long abstractive summarization task. Despite faster decoding speeds compared to autoregressive methods, previous discrete diffusion models failed on the abstractive summarization task due to the incompatibility between the backbone architectures and the random noising process. To overcome these challenges, we introduce a novel semantic-aware noising process that enables Transformer backbones to handle long sequences effectively. Additionally, we propose CrossMamba, an adaptation of the Mamba model to the encoder-decoder paradigm, which integrates seamlessly with the random absorbing noising process. Our approaches outperform existing discrete diffusion models on three benchmark summarization datasets: Gigaword, CNN/DailyMail, and Arxiv, while also achieving much faster inference speed compared to autoregressive models.
%R 10.18653/v1/2025.findings-naacl.352
%U https://aclanthology.org/2025.findings-naacl.352/
%U https://doi.org/10.18653/v1/2025.findings-naacl.352
%P 6278-6290
Markdown (Informal)
[Discrete Diffusion Language Model for Efficient Text Summarization](https://aclanthology.org/2025.findings-naacl.352/) (Dat et al., Findings 2025)
ACL