@inproceedings{lee-etal-2026-tagquant,
title = "{TAGQ}uant: Token-Aware Clustering for Group-Wise Quantization",
author = "Lee, Jaeseong and
Hwang, Seung-won and
Qiao, Aurick and
Yao, Zhewei and
He, Yuxiong",
editor = {Matusevych, Yevgen and
Eryi{\u{g}}it, G{\"u}l{\c{s}}en and
Aletras, Nikolaos},
booktitle = "Proceedings of the 19th Conference of the {E}uropean Chapter of the {A}ssociation for {C}omputational {L}inguistics (Volume 5: Industry Track)",
month = mar,
year = "2026",
address = "Rabat, Morocco",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.eacl-industry.18/",
pages = "253--262",
ISBN = "979-8-89176-384-5",
abstract = "Grouping, e.g., grouping channels, which is widely used in current integer-based quantization, has become essential for the emerging MXFP4 format. Ideally, each group should contain channels with similar quantization scales. To guide such groups, existing work clusters the channels using scalar proxy, ignoring the token dimension, which we find suboptimal. In this paper, we propose TAGQuant, a simple yet powerful enhancement for such ``group-wise'' quantization. By strategically shuffling channels to group those with similar token-wise activation distributions, TAGQuant ensures better clustering of large- and small-range values. This shuffle operation is hardware-efficient, and seamlessly integrated into the quantization process with only 0.01x latency overhead. TAGQuant reduces relative GSM8K error in both INT4 and MXFP4 formats, by up to 86{\%} in Llama-3.1-8B-Instruct compared to baselines, validating the effectiveness of our channel shuffling approach for group-wise quantization. Code is publicly available."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="lee-etal-2026-tagquant">
<titleInfo>
<title>TAGQuant: Token-Aware Clustering for Group-Wise Quantization</title>
</titleInfo>
<name type="personal">
<namePart type="given">Jaeseong</namePart>
<namePart type="family">Lee</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Seung-won</namePart>
<namePart type="family">Hwang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Aurick</namePart>
<namePart type="family">Qiao</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Zhewei</namePart>
<namePart type="family">Yao</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yuxiong</namePart>
<namePart type="family">He</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 5: Industry Track)</title>
</titleInfo>
<name type="personal">
<namePart type="given">Yevgen</namePart>
<namePart type="family">Matusevych</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Gülşen</namePart>
<namePart type="family">Eryiğit</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Nikolaos</namePart>
<namePart type="family">Aletras</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">979-8-89176-384-5</identifier>
</relatedItem>
<abstract>Grouping, e.g., grouping channels, which is widely used in current integer-based quantization, has become essential for the emerging MXFP4 format. Ideally, each group should contain channels with similar quantization scales. To guide such groups, existing work clusters the channels using scalar proxy, ignoring the token dimension, which we find suboptimal. In this paper, we propose TAGQuant, a simple yet powerful enhancement for such “group-wise” quantization. By strategically shuffling channels to group those with similar token-wise activation distributions, TAGQuant ensures better clustering of large- and small-range values. This shuffle operation is hardware-efficient, and seamlessly integrated into the quantization process with only 0.01x latency overhead. TAGQuant reduces relative GSM8K error in both INT4 and MXFP4 formats, by up to 86% in Llama-3.1-8B-Instruct compared to baselines, validating the effectiveness of our channel shuffling approach for group-wise quantization. Code is publicly available.</abstract>
<identifier type="citekey">lee-etal-2026-tagquant</identifier>
<location>
<url>https://aclanthology.org/2026.eacl-industry.18/</url>
</location>
<part>
<date>2026-03</date>
<extent unit="page">
<start>253</start>
<end>262</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T TAGQuant: Token-Aware Clustering for Group-Wise Quantization
%A Lee, Jaeseong
%A Hwang, Seung-won
%A Qiao, Aurick
%A Yao, Zhewei
%A He, Yuxiong
%Y Matusevych, Yevgen
%Y Eryiğit, Gülşen
%Y Aletras, Nikolaos
%S Proceedings of the 19th Conference of the European Chapter of the Association for Computational Linguistics (Volume 5: Industry Track)
%D 2026
%8 March
%I Association for Computational Linguistics
%C Rabat, Morocco
%@ 979-8-89176-384-5
%F lee-etal-2026-tagquant
%X Grouping, e.g., grouping channels, which is widely used in current integer-based quantization, has become essential for the emerging MXFP4 format. Ideally, each group should contain channels with similar quantization scales. To guide such groups, existing work clusters the channels using scalar proxy, ignoring the token dimension, which we find suboptimal. In this paper, we propose TAGQuant, a simple yet powerful enhancement for such “group-wise” quantization. By strategically shuffling channels to group those with similar token-wise activation distributions, TAGQuant ensures better clustering of large- and small-range values. This shuffle operation is hardware-efficient, and seamlessly integrated into the quantization process with only 0.01x latency overhead. TAGQuant reduces relative GSM8K error in both INT4 and MXFP4 formats, by up to 86% in Llama-3.1-8B-Instruct compared to baselines, validating the effectiveness of our channel shuffling approach for group-wise quantization. Code is publicly available.
%U https://aclanthology.org/2026.eacl-industry.18/
%P 253-262
Markdown (Informal)
[TAGQuant: Token-Aware Clustering for Group-Wise Quantization](https://aclanthology.org/2026.eacl-industry.18/) (Lee et al., EACL 2026)
ACL
- Jaeseong Lee, Seung-won Hwang, Aurick Qiao, Zhewei Yao, and Yuxiong He. 2026. TAGQuant: Token-Aware Clustering for Group-Wise Quantization. In Proceedings of the 19th Conference of the European Chapter of the Association for Computational Linguistics (Volume 5: Industry Track), pages 253–262, Rabat, Morocco. Association for Computational Linguistics.