@inproceedings{dong-etal-2026-multi,
title = "Multi-Scale Model Compression via Nested Matrix Learning",
author = "Dong, Xiangjue and
Anantharaman, Aditya and
Pugaliya, Hemant and
Zhong, Kai",
editor = "Piperidis, Stelios and
Bel, N{\'u}ria and
van den Heuvel, Henk and
Ide, Nancy and
Krek, Simon and
Toral, Antonio",
booktitle = "Proceedings of the Fifteenth Language Resources and Evaluation Conference",
month = may,
year = "2026",
address = "Palma de Mallorca, Spain",
publisher = "ELRA Language Resource Association",
url = "https://aclanthology.org/2026.lrec-1.196/",
doi = "10.63317/5o97c4anqod5",
pages = "2501--2511",
abstract = "Large language models (LLMs) have been widely deployed and have achieved remarkable success in downstream tasks. However, their high latency continues to pose challenges for real-time applications that require fast inference, and the need to train and deploy distinct models for different hardware constraints increases both financial and computational costs. To address this, we propose Nested Matrix Learning (NML), a method that trains a single, flexible model capable of generating multiple high-performing student models of varying sizes. This is achieved by simultaneously optimizing a pre-trained teacher model and its nested sub-models in a single training process, without sacrificing the teacher{'}s performance. NML provides a flexible and scalable solution, allowing models to adapt to different computational budgets. Our extensive experiments show that student models produced by NML, which can be up to 10x smaller than the full-size model, can be directly deployed for efficient inference or serve as superior initialization points for further fine-tuning in downstream tasks. By preserving the performance of the teacher model while delivering compact and efficient student models of various sizes, NML enhances the usability and adaptability of LLMs in real-world scenarios."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="dong-etal-2026-multi">
<titleInfo>
<title>Multi-Scale Model Compression via Nested Matrix Learning</title>
</titleInfo>
<name type="personal">
<namePart type="given">Xiangjue</namePart>
<namePart type="family">Dong</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Aditya</namePart>
<namePart type="family">Anantharaman</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Hemant</namePart>
<namePart type="family">Pugaliya</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Kai</namePart>
<namePart type="family">Zhong</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-05</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the Fifteenth Language Resources and Evaluation Conference</title>
</titleInfo>
<name type="personal">
<namePart type="given">Stelios</namePart>
<namePart type="family">Piperidis</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Núria</namePart>
<namePart type="family">Bel</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Henk</namePart>
<namePart type="family">van den Heuvel</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Nancy</namePart>
<namePart type="family">Ide</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Simon</namePart>
<namePart type="family">Krek</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Antonio</namePart>
<namePart type="family">Toral</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>ELRA Language Resource Association</publisher>
<place>
<placeTerm type="text">Palma de Mallorca, Spain</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>Large language models (LLMs) have been widely deployed and have achieved remarkable success in downstream tasks. However, their high latency continues to pose challenges for real-time applications that require fast inference, and the need to train and deploy distinct models for different hardware constraints increases both financial and computational costs. To address this, we propose Nested Matrix Learning (NML), a method that trains a single, flexible model capable of generating multiple high-performing student models of varying sizes. This is achieved by simultaneously optimizing a pre-trained teacher model and its nested sub-models in a single training process, without sacrificing the teacher’s performance. NML provides a flexible and scalable solution, allowing models to adapt to different computational budgets. Our extensive experiments show that student models produced by NML, which can be up to 10x smaller than the full-size model, can be directly deployed for efficient inference or serve as superior initialization points for further fine-tuning in downstream tasks. By preserving the performance of the teacher model while delivering compact and efficient student models of various sizes, NML enhances the usability and adaptability of LLMs in real-world scenarios.</abstract>
<identifier type="citekey">dong-etal-2026-multi</identifier>
<identifier type="doi">10.63317/5o97c4anqod5</identifier>
<location>
<url>https://aclanthology.org/2026.lrec-1.196/</url>
</location>
<part>
<date>2026-05</date>
<extent unit="page">
<start>2501</start>
<end>2511</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Multi-Scale Model Compression via Nested Matrix Learning
%A Dong, Xiangjue
%A Anantharaman, Aditya
%A Pugaliya, Hemant
%A Zhong, Kai
%Y Piperidis, Stelios
%Y Bel, Núria
%Y van den Heuvel, Henk
%Y Ide, Nancy
%Y Krek, Simon
%Y Toral, Antonio
%S Proceedings of the Fifteenth Language Resources and Evaluation Conference
%D 2026
%8 May
%I ELRA Language Resource Association
%C Palma de Mallorca, Spain
%F dong-etal-2026-multi
%X Large language models (LLMs) have been widely deployed and have achieved remarkable success in downstream tasks. However, their high latency continues to pose challenges for real-time applications that require fast inference, and the need to train and deploy distinct models for different hardware constraints increases both financial and computational costs. To address this, we propose Nested Matrix Learning (NML), a method that trains a single, flexible model capable of generating multiple high-performing student models of varying sizes. This is achieved by simultaneously optimizing a pre-trained teacher model and its nested sub-models in a single training process, without sacrificing the teacher’s performance. NML provides a flexible and scalable solution, allowing models to adapt to different computational budgets. Our extensive experiments show that student models produced by NML, which can be up to 10x smaller than the full-size model, can be directly deployed for efficient inference or serve as superior initialization points for further fine-tuning in downstream tasks. By preserving the performance of the teacher model while delivering compact and efficient student models of various sizes, NML enhances the usability and adaptability of LLMs in real-world scenarios.
%R 10.63317/5o97c4anqod5
%U https://aclanthology.org/2026.lrec-1.196/
%U https://doi.org/10.63317/5o97c4anqod5
%P 2501-2511
Markdown (Informal)
[Multi-Scale Model Compression via Nested Matrix Learning](https://aclanthology.org/2026.lrec-1.196/) (Dong et al., LREC 2026)
ACL
- Xiangjue Dong, Aditya Anantharaman, Hemant Pugaliya, and Kai Zhong. 2026. Multi-Scale Model Compression via Nested Matrix Learning. In Proceedings of the Fifteenth Language Resources and Evaluation Conference, pages 2501–2511, Palma de Mallorca, Spain. ELRA Language Resource Association.