@inproceedings{zong-etal-2026-l2dir,
title = "{L}2{D}ir: Integrating $L\_2$-Norm and Directional Alignment for Unsupervised Contrastive Representation Learning in Multimodal Retrieval",
author = "Zong, Tianyu and
Dai, Rui and
Yi, Hongzhu and
Wang, Yuanxiang and
Zhang, Zhenghao and
Guan, Zhenyu and
Yang, Yujia and
Shi, Bingkang and
Ding, Yueyang and
Chu, Xiangxiang and
Liu, Kaikui and
Xu, Jungang",
editor = "Liakata, Maria and
Moreira, Viviane P. and
Zhang, Jiajun and
Jurgens, David",
booktitle = "Proceedings of the 64th Annual Meeting of the {A}ssociation for {C}omputational {L}inguistics (Volume 1: Long Papers)",
month = jul,
year = "2026",
address = "San Diego, California, United States",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.acl-long.1604/",
pages = "34717--34740",
ISBN = "979-8-89176-390-6",
abstract = "Multimodal representation learning primarily relies on contrastive objectives such as InfoNCE to align diverse modalities. However, these methods focus almost exclusively on directional alignment and often neglect the intrinsic role of embedding magnitudes (L2-norm) in the contrastive process. To bridge this gap, we propose L2Dir, a plug-and-play framework designed to optimize L2-norm alignment and Directional consistency jointly. As a highly efficient solution, L2Dir doesn{'}t require extra data, distillation, or external supervision. It can be integrated seamlessly into existing pipelines by employing a lightweight MLP to reconstruct magnitudes from frozen backbone features. Extensive evaluations across 95 tasks using UniIR and VLM2Vec-V2 frameworks demonstrate that L2Dir yields consistent and significant performance gains over established baselines across various backbones and scales, proving that explicit magnitude modeling is a versatile and potent strategy for refining unsupervised multimodal representations. The source code for L2Dir in VLM2Vec-V2 is available in the supplementary materials."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="zong-etal-2026-l2dir">
<titleInfo>
<title>L2Dir: Integrating L_2-Norm and Directional Alignment for Unsupervised Contrastive Representation Learning in Multimodal Retrieval</title>
</titleInfo>
<name type="personal">
<namePart type="given">Tianyu</namePart>
<namePart type="family">Zong</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Rui</namePart>
<namePart type="family">Dai</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Hongzhu</namePart>
<namePart type="family">Yi</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yuanxiang</namePart>
<namePart type="family">Wang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Zhenghao</namePart>
<namePart type="family">Zhang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Zhenyu</namePart>
<namePart type="family">Guan</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yujia</namePart>
<namePart type="family">Yang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Bingkang</namePart>
<namePart type="family">Shi</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yueyang</namePart>
<namePart type="family">Ding</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Xiangxiang</namePart>
<namePart type="family">Chu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Kaikui</namePart>
<namePart type="family">Liu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jungang</namePart>
<namePart type="family">Xu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-07</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)</title>
</titleInfo>
<name type="personal">
<namePart type="given">Maria</namePart>
<namePart type="family">Liakata</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Viviane</namePart>
<namePart type="given">P</namePart>
<namePart type="family">Moreira</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jiajun</namePart>
<namePart type="family">Zhang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">David</namePart>
<namePart type="family">Jurgens</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">San Diego, California, United States</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-390-6</identifier>
</relatedItem>
<abstract>Multimodal representation learning primarily relies on contrastive objectives such as InfoNCE to align diverse modalities. However, these methods focus almost exclusively on directional alignment and often neglect the intrinsic role of embedding magnitudes (L2-norm) in the contrastive process. To bridge this gap, we propose L2Dir, a plug-and-play framework designed to optimize L2-norm alignment and Directional consistency jointly. As a highly efficient solution, L2Dir doesn’t require extra data, distillation, or external supervision. It can be integrated seamlessly into existing pipelines by employing a lightweight MLP to reconstruct magnitudes from frozen backbone features. Extensive evaluations across 95 tasks using UniIR and VLM2Vec-V2 frameworks demonstrate that L2Dir yields consistent and significant performance gains over established baselines across various backbones and scales, proving that explicit magnitude modeling is a versatile and potent strategy for refining unsupervised multimodal representations. The source code for L2Dir in VLM2Vec-V2 is available in the supplementary materials.</abstract>
<identifier type="citekey">zong-etal-2026-l2dir</identifier>
<location>
<url>https://aclanthology.org/2026.acl-long.1604/</url>
</location>
<part>
<date>2026-07</date>
<extent unit="page">
<start>34717</start>
<end>34740</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T L2Dir: Integrating L_2-Norm and Directional Alignment for Unsupervised Contrastive Representation Learning in Multimodal Retrieval
%A Zong, Tianyu
%A Dai, Rui
%A Yi, Hongzhu
%A Wang, Yuanxiang
%A Zhang, Zhenghao
%A Guan, Zhenyu
%A Yang, Yujia
%A Shi, Bingkang
%A Ding, Yueyang
%A Chu, Xiangxiang
%A Liu, Kaikui
%A Xu, Jungang
%Y Liakata, Maria
%Y Moreira, Viviane P.
%Y Zhang, Jiajun
%Y Jurgens, David
%S Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)
%D 2026
%8 July
%I Association for Computational Linguistics
%C San Diego, California, United States
%@ 979-8-89176-390-6
%F zong-etal-2026-l2dir
%X Multimodal representation learning primarily relies on contrastive objectives such as InfoNCE to align diverse modalities. However, these methods focus almost exclusively on directional alignment and often neglect the intrinsic role of embedding magnitudes (L2-norm) in the contrastive process. To bridge this gap, we propose L2Dir, a plug-and-play framework designed to optimize L2-norm alignment and Directional consistency jointly. As a highly efficient solution, L2Dir doesn’t require extra data, distillation, or external supervision. It can be integrated seamlessly into existing pipelines by employing a lightweight MLP to reconstruct magnitudes from frozen backbone features. Extensive evaluations across 95 tasks using UniIR and VLM2Vec-V2 frameworks demonstrate that L2Dir yields consistent and significant performance gains over established baselines across various backbones and scales, proving that explicit magnitude modeling is a versatile and potent strategy for refining unsupervised multimodal representations. The source code for L2Dir in VLM2Vec-V2 is available in the supplementary materials.
%U https://aclanthology.org/2026.acl-long.1604/
%P 34717-34740
Markdown (Informal)
[L2Dir: Integrating L_2-Norm and Directional Alignment for Unsupervised Contrastive Representation Learning in Multimodal Retrieval](https://aclanthology.org/2026.acl-long.1604/) (Zong et al., ACL 2026)
ACL
- Tianyu Zong, Rui Dai, Hongzhu Yi, Yuanxiang Wang, Zhenghao Zhang, Zhenyu Guan, Yujia Yang, Bingkang Shi, Yueyang Ding, Xiangxiang Chu, Kaikui Liu, and Jungang Xu. 2026. L2Dir: Integrating L_2-Norm and Directional Alignment for Unsupervised Contrastive Representation Learning in Multimodal Retrieval. In Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 34717–34740, San Diego, California, United States. Association for Computational Linguistics.