@inproceedings{xiao-zhong-2026-beyond,
title = "Beyond Redundancy: Information-Conserving Semantic Packaging for Vision-Language Models",
author = "Xiao, Xin and
Zhong, Jiang",
editor = "Lee, Hung-Yi and
Dev, Sunipa and
Li, Jessy and
Elliott, Desmond",
booktitle = "Findings of the {A}ssociation for {C}omputational {L}inguistics: {EMNLP} 2026",
month = oct,
year = "2026",
address = "Budapest, Hungary",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.findings-emnlp.848/",
pages = "17970--17980",
ISBN = "979-8-89176-462-0",
abstract = "Vision-Language Models (VLMs) achieve strong multimodal understanding but suffer from high inference costs due to redundant vision tokens. Existing hard-pruning methods discard low-scoring tokens directly, inevitably sacrificing fine-grained visual context. We propose SPack, a training-free token compression method that reframes pruning as information-energy redistribution rather than deletion. Instead of discarding pruned tokens, SPack transfers their semantic mass to retained tokens, enforcing a global principle we term Information Energy Conservation (IEC): the total information contained in redundant tokens is strictly conserved and fully transferred as information gain to the retained components. To realize this conservation, we introduce Optimal Transport (OT) theory, which treats removed energy as mass density and retained tokens as receiver slots. Solving the optimal transport plan under the Wasserstein distance yields a minimum-cost mapping that precisely transfers geometric structure and higher-order moments according to semantic affinity. Concretely, we estimate token importance via relation attention, semantic affinity, and value-space energy, then adaptively route pruned tokens to multiple retained tokens using OT, producing hyper-dense tokens that preserve both local semantics and global context. Extensive experiments demonstrate that SPack consistently achieves a superior accuracy-efficiency trade-off compared to state-of-the-art methods. The code will be released upon acceptance."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="xiao-zhong-2026-beyond">
<titleInfo>
<title>Beyond Redundancy: Information-Conserving Semantic Packaging for Vision-Language Models</title>
</titleInfo>
<name type="personal">
<namePart type="given">Xin</namePart>
<namePart type="family">Xiao</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jiang</namePart>
<namePart type="family">Zhong</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-10</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Findings of the Association for Computational Linguistics: EMNLP 2026</title>
</titleInfo>
<name type="personal">
<namePart type="given">Hung-Yi</namePart>
<namePart type="family">Lee</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Sunipa</namePart>
<namePart type="family">Dev</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jessy</namePart>
<namePart type="family">Li</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Desmond</namePart>
<namePart type="family">Elliott</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Budapest, Hungary</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-462-0</identifier>
</relatedItem>
<abstract>Vision-Language Models (VLMs) achieve strong multimodal understanding but suffer from high inference costs due to redundant vision tokens. Existing hard-pruning methods discard low-scoring tokens directly, inevitably sacrificing fine-grained visual context. We propose SPack, a training-free token compression method that reframes pruning as information-energy redistribution rather than deletion. Instead of discarding pruned tokens, SPack transfers their semantic mass to retained tokens, enforcing a global principle we term Information Energy Conservation (IEC): the total information contained in redundant tokens is strictly conserved and fully transferred as information gain to the retained components. To realize this conservation, we introduce Optimal Transport (OT) theory, which treats removed energy as mass density and retained tokens as receiver slots. Solving the optimal transport plan under the Wasserstein distance yields a minimum-cost mapping that precisely transfers geometric structure and higher-order moments according to semantic affinity. Concretely, we estimate token importance via relation attention, semantic affinity, and value-space energy, then adaptively route pruned tokens to multiple retained tokens using OT, producing hyper-dense tokens that preserve both local semantics and global context. Extensive experiments demonstrate that SPack consistently achieves a superior accuracy-efficiency trade-off compared to state-of-the-art methods. The code will be released upon acceptance.</abstract>
<identifier type="citekey">xiao-zhong-2026-beyond</identifier>
<location>
<url>https://aclanthology.org/2026.findings-emnlp.848/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>17970</start>
<end>17980</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Beyond Redundancy: Information-Conserving Semantic Packaging for Vision-Language Models
%A Xiao, Xin
%A Zhong, Jiang
%Y Lee, Hung-Yi
%Y Dev, Sunipa
%Y Li, Jessy
%Y Elliott, Desmond
%S Findings of the Association for Computational Linguistics: EMNLP 2026
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-462-0
%F xiao-zhong-2026-beyond
%X Vision-Language Models (VLMs) achieve strong multimodal understanding but suffer from high inference costs due to redundant vision tokens. Existing hard-pruning methods discard low-scoring tokens directly, inevitably sacrificing fine-grained visual context. We propose SPack, a training-free token compression method that reframes pruning as information-energy redistribution rather than deletion. Instead of discarding pruned tokens, SPack transfers their semantic mass to retained tokens, enforcing a global principle we term Information Energy Conservation (IEC): the total information contained in redundant tokens is strictly conserved and fully transferred as information gain to the retained components. To realize this conservation, we introduce Optimal Transport (OT) theory, which treats removed energy as mass density and retained tokens as receiver slots. Solving the optimal transport plan under the Wasserstein distance yields a minimum-cost mapping that precisely transfers geometric structure and higher-order moments according to semantic affinity. Concretely, we estimate token importance via relation attention, semantic affinity, and value-space energy, then adaptively route pruned tokens to multiple retained tokens using OT, producing hyper-dense tokens that preserve both local semantics and global context. Extensive experiments demonstrate that SPack consistently achieves a superior accuracy-efficiency trade-off compared to state-of-the-art methods. The code will be released upon acceptance.
%U https://aclanthology.org/2026.findings-emnlp.848/
%P 17970-17980
Markdown (Informal)
[Beyond Redundancy: Information-Conserving Semantic Packaging for Vision-Language Models](https://aclanthology.org/2026.findings-emnlp.848/) (Xiao & Zhong, Findings 2026)
ACL