@inproceedings{williams-etal-2026-compressing,
title = "Compressing Language Models for Specialized Domains",
author = "Williams, Miles and
Chrysostomou, George and
Jeronymo, Vitor Amancio and
Aletras, Nikolaos",
editor = "Demberg, Vera and
Inui, Kentaro and
Marquez, Llu{\'i}s",
booktitle = "Proceedings of the 19th Conference of the {E}uropean Chapter of the {A}ssociation for {C}omputational {L}inguistics (Volume 1: Long Papers)",
month = mar,
year = "2026",
address = "Rabat, Morocco",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.eacl-long.347/",
pages = "7393--7415",
ISBN = "979-8-89176-380-7",
abstract = "Language models (LMs) excel at tasks across diverse domains, yet require substantial computational resources during inference. Compression techniques such as pruning and quantization offer a practical path towards efficient LM deployment, exemplified by their ability to preserve performance on general-purpose benchmarks. However, general-purpose LM compression methods can negatively affect performance in specialized domains (e.g. biomedical or legal). Recent work has sought to address this issue, but requires a computationally expensive full-parameter fine-tuning pipeline. To this end, we propose MixCal, a novel calibration method designed to improve the in-domain performance of compressed LMs in a post-training setting. Through extensive experimentation, we demonstrate that MixCal substantially outperforms existing approaches on domain-specific tasks while preserving general performance. Notably, these performance gains are achieved while also reducing the computational cost of LM compression."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="williams-etal-2026-compressing">
<titleInfo>
<title>Compressing Language Models for Specialized Domains</title>
</titleInfo>
<name type="personal">
<namePart type="given">Miles</namePart>
<namePart type="family">Williams</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">George</namePart>
<namePart type="family">Chrysostomou</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Vitor</namePart>
<namePart type="given">Amancio</namePart>
<namePart type="family">Jeronymo</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Nikolaos</namePart>
<namePart type="family">Aletras</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-03</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 19th Conference of the European Chapter of the Association for Computational Linguistics (Volume 1: Long Papers)</title>
</titleInfo>
<name type="personal">
<namePart type="given">Vera</namePart>
<namePart type="family">Demberg</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Kentaro</namePart>
<namePart type="family">Inui</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Lluís</namePart>
<namePart type="family">Marquez</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Rabat, Morocco</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-380-7</identifier>
</relatedItem>
<abstract>Language models (LMs) excel at tasks across diverse domains, yet require substantial computational resources during inference. Compression techniques such as pruning and quantization offer a practical path towards efficient LM deployment, exemplified by their ability to preserve performance on general-purpose benchmarks. However, general-purpose LM compression methods can negatively affect performance in specialized domains (e.g. biomedical or legal). Recent work has sought to address this issue, but requires a computationally expensive full-parameter fine-tuning pipeline. To this end, we propose MixCal, a novel calibration method designed to improve the in-domain performance of compressed LMs in a post-training setting. Through extensive experimentation, we demonstrate that MixCal substantially outperforms existing approaches on domain-specific tasks while preserving general performance. Notably, these performance gains are achieved while also reducing the computational cost of LM compression.</abstract>
<identifier type="citekey">williams-etal-2026-compressing</identifier>
<location>
<url>https://aclanthology.org/2026.eacl-long.347/</url>
</location>
<part>
<date>2026-03</date>
<extent unit="page">
<start>7393</start>
<end>7415</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Compressing Language Models for Specialized Domains
%A Williams, Miles
%A Chrysostomou, George
%A Jeronymo, Vitor Amancio
%A Aletras, Nikolaos
%Y Demberg, Vera
%Y Inui, Kentaro
%Y Marquez, Lluís
%S Proceedings of the 19th Conference of the European Chapter of the Association for Computational Linguistics (Volume 1: Long Papers)
%D 2026
%8 March
%I Association for Computational Linguistics
%C Rabat, Morocco
%@ 979-8-89176-380-7
%F williams-etal-2026-compressing
%X Language models (LMs) excel at tasks across diverse domains, yet require substantial computational resources during inference. Compression techniques such as pruning and quantization offer a practical path towards efficient LM deployment, exemplified by their ability to preserve performance on general-purpose benchmarks. However, general-purpose LM compression methods can negatively affect performance in specialized domains (e.g. biomedical or legal). Recent work has sought to address this issue, but requires a computationally expensive full-parameter fine-tuning pipeline. To this end, we propose MixCal, a novel calibration method designed to improve the in-domain performance of compressed LMs in a post-training setting. Through extensive experimentation, we demonstrate that MixCal substantially outperforms existing approaches on domain-specific tasks while preserving general performance. Notably, these performance gains are achieved while also reducing the computational cost of LM compression.
%U https://aclanthology.org/2026.eacl-long.347/
%P 7393-7415
Markdown (Informal)
[Compressing Language Models for Specialized Domains](https://aclanthology.org/2026.eacl-long.347/) (Williams et al., EACL 2026)
ACL
- Miles Williams, George Chrysostomou, Vitor Amancio Jeronymo, and Nikolaos Aletras. 2026. Compressing Language Models for Specialized Domains. In Proceedings of the 19th Conference of the European Chapter of the Association for Computational Linguistics (Volume 1: Long Papers), pages 7393–7415, Rabat, Morocco. Association for Computational Linguistics.