@inproceedings{inkiriwang-etal-2025-really,
title = "Do We Really Need All Those Dimensions? An Intrinsic Evaluation Framework for Compressed Embeddings",
author = {Inkiriwang, Nathan and
B{\"o}l{\"u}c{\"u}, Necva and
Tarr, Garth and
Rybinski, Maciej},
editor = "Christodoulopoulos, Christos and
Chakraborty, Tanmoy and
Rose, Carolyn and
Peng, Violet",
booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2025",
month = nov,
year = "2025",
address = "Suzhou, China",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2025.findings-emnlp.717/",
pages = "13305--13323",
ISBN = "979-8-89176-335-7",
abstract = "High-dimensional text embeddings are foundational to modern NLP but costly to store and use. While embedding compression addresses these challenges, selecting the best compression method remains difficult. Existing evaluation methods for compressed embeddings are either expensive or too simplistic. We introduce a comprehensive intrinsic evaluation framework featuring a suite of task-agnostic metrics that together provide a reliable proxy for downstream performance. A key contribution is $\operatorname{EOS}_k$, a novel spectral fidelity measure specifically designed to be robust to embedding anisotropy. Through extensive experiments on diverse embeddings across four downstream tasks, we demonstrate that our intrinsic metrics reliably predict extrinsic performance and reveal how different embedding architectures depend on distinct geometric properties. Our framework provides a practical, efficient, and interpretable alternative to standard evaluations for compressed embeddings."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="inkiriwang-etal-2025-really">
<titleInfo>
<title>Do We Really Need All Those Dimensions? An Intrinsic Evaluation Framework for Compressed Embeddings</title>
</titleInfo>
<name type="personal">
<namePart type="given">Nathan</namePart>
<namePart type="family">Inkiriwang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Necva</namePart>
<namePart type="family">Bölücü</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Garth</namePart>
<namePart type="family">Tarr</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Maciej</namePart>
<namePart type="family">Rybinski</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2025-11</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Findings of the Association for Computational Linguistics: EMNLP 2025</title>
</titleInfo>
<name type="personal">
<namePart type="given">Christos</namePart>
<namePart type="family">Christodoulopoulos</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Tanmoy</namePart>
<namePart type="family">Chakraborty</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Carolyn</namePart>
<namePart type="family">Rose</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Violet</namePart>
<namePart type="family">Peng</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Suzhou, China</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-335-7</identifier>
</relatedItem>
<abstract>High-dimensional text embeddings are foundational to modern NLP but costly to store and use. While embedding compression addresses these challenges, selecting the best compression method remains difficult. Existing evaluation methods for compressed embeddings are either expensive or too simplistic. We introduce a comprehensive intrinsic evaluation framework featuring a suite of task-agnostic metrics that together provide a reliable proxy for downstream performance. A key contribution is øperatornameEOS_k, a novel spectral fidelity measure specifically designed to be robust to embedding anisotropy. Through extensive experiments on diverse embeddings across four downstream tasks, we demonstrate that our intrinsic metrics reliably predict extrinsic performance and reveal how different embedding architectures depend on distinct geometric properties. Our framework provides a practical, efficient, and interpretable alternative to standard evaluations for compressed embeddings.</abstract>
<identifier type="citekey">inkiriwang-etal-2025-really</identifier>
<location>
<url>https://aclanthology.org/2025.findings-emnlp.717/</url>
</location>
<part>
<date>2025-11</date>
<extent unit="page">
<start>13305</start>
<end>13323</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Do We Really Need All Those Dimensions? An Intrinsic Evaluation Framework for Compressed Embeddings
%A Inkiriwang, Nathan
%A Bölücü, Necva
%A Tarr, Garth
%A Rybinski, Maciej
%Y Christodoulopoulos, Christos
%Y Chakraborty, Tanmoy
%Y Rose, Carolyn
%Y Peng, Violet
%S Findings of the Association for Computational Linguistics: EMNLP 2025
%D 2025
%8 November
%I Association for Computational Linguistics
%C Suzhou, China
%@ 979-8-89176-335-7
%F inkiriwang-etal-2025-really
%X High-dimensional text embeddings are foundational to modern NLP but costly to store and use. While embedding compression addresses these challenges, selecting the best compression method remains difficult. Existing evaluation methods for compressed embeddings are either expensive or too simplistic. We introduce a comprehensive intrinsic evaluation framework featuring a suite of task-agnostic metrics that together provide a reliable proxy for downstream performance. A key contribution is øperatornameEOS_k, a novel spectral fidelity measure specifically designed to be robust to embedding anisotropy. Through extensive experiments on diverse embeddings across four downstream tasks, we demonstrate that our intrinsic metrics reliably predict extrinsic performance and reveal how different embedding architectures depend on distinct geometric properties. Our framework provides a practical, efficient, and interpretable alternative to standard evaluations for compressed embeddings.
%U https://aclanthology.org/2025.findings-emnlp.717/
%P 13305-13323
Markdown (Informal)
[Do We Really Need All Those Dimensions? An Intrinsic Evaluation Framework for Compressed Embeddings](https://aclanthology.org/2025.findings-emnlp.717/) (Inkiriwang et al., Findings 2025)
ACL