<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="lizesheng-etal-2026-grasp">
    <titleInfo>
        <title>GraSP-VL: Length as a Semantic Granularity Interface for Vision-Language Representations</title>
    </titleInfo>
    <name>
        <namePart>Lizesheng</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Pan</namePart>
        <namePart type="family">Chengchang</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Honggang</namePart>
        <namePart type="family">Qi</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>Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing</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">979-8-89176-459-0</identifier>
    </relatedItem>
    <abstract>Frozen vision-language models already encode object, attribute, relation, and caption information, but a single vector exposes all of these signals through the same access pattern. We ask whether vector length can become a semantic access control: short prefixes answer coarse queries, while longer prefixes add finer distinctions. GraSP-VL learns one shared near-orthogonal transform for image and text embeddings and assigns a fixed, configurable prefix contract. At inference, the transform is applied once and ordinary cosine retrieval uses the prefix required by the task. On a 20,147-example COCO/Flickr30K pool, GraSP-VL reaches a staircase score of 53.01 and hard-negative selectivity of 89.76 with full-space drift below 10⁻6. Full-dimensional ImageNet-V2 classification and Flickr30K/COCO Karpathy retrieval are unchanged, while held-out human-caption retrieval improves at shorter prefixes. These results show that length can reorganize access to frozen VLM semantics rather than merely compressing them. Code and project page are available at https://lizesheng13.github.io/Grasp-VL/.</abstract>
    <identifier type="citekey">lizesheng-etal-2026-grasp</identifier>
    <location>
        <url>https://aclanthology.org/2026.emnlp-main.344/</url>
    </location>
    <part>
        <date>2026-10</date>
        <extent unit="page">
            <start>7919</start>
            <end>7935</end>
        </extent>
    </part>
</mods>
</modsCollection>

%0 Conference Proceedings
%T GraSP-VL: Length as a Semantic Granularity Interface for Vision-Language Representations
%A Chengchang, Pan
%A Qi, Honggang
%Y Lee, Hung-Yi
%Y Dev, Sunipa
%Y Li, Jessy
%Y Elliott, Desmond
%A Lizesheng
%S Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-459-0
%F lizesheng-etal-2026-grasp
%X Frozen vision-language models already encode object, attribute, relation, and caption information, but a single vector exposes all of these signals through the same access pattern. We ask whether vector length can become a semantic access control: short prefixes answer coarse queries, while longer prefixes add finer distinctions. GraSP-VL learns one shared near-orthogonal transform for image and text embeddings and assigns a fixed, configurable prefix contract. At inference, the transform is applied once and ordinary cosine retrieval uses the prefix required by the task. On a 20,147-example COCO/Flickr30K pool, GraSP-VL reaches a staircase score of 53.01 and hard-negative selectivity of 89.76 with full-space drift below 10⁻6. Full-dimensional ImageNet-V2 classification and Flickr30K/COCO Karpathy retrieval are unchanged, while held-out human-caption retrieval improves at shorter prefixes. These results show that length can reorganize access to frozen VLM semantics rather than merely compressing them. Code and project page are available at https://lizesheng13.github.io/Grasp-VL/.
%U https://aclanthology.org/2026.emnlp-main.344/
%P 7919-7935

Markdown (Informal)

[GraSP-VL: Length as a Semantic Granularity Interface for Vision-Language Representations](https://aclanthology.org/2026.emnlp-main.344/) (Lizesheng et al., EMNLP 2026)

• GraSP-VL: Length as a Semantic Granularity Interface for Vision-Language Representations (Lizesheng et al., EMNLP 2026)

ACL

• Lizesheng, Pan Chengchang, and Honggang Qi. 2026. GraSP-VL: Length as a Semantic Granularity Interface for Vision-Language Representations. In Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing, pages 7919–7935, Budapest, Hungary. Association for Computational Linguistics.