@inproceedings{cherepanov-burlova-2026-occupy,
title = "What Should Occupy the Context Budget? Latent Compression versus Budget-Matched Truncation in Document {QA}",
author = "Cherepanov, Artem and
Burlova, Albina",
editor = "T.Y.S.S., Santosh and
Agarwal, Amit and
Gupta, Vivek and
Kementchedjhieva, Yova and
Garimella, Aparna and
Hou, Yufang",
booktitle = "The Proceedings of Workshop on Document Intelligence and Understanding",
month = oct,
year = "2026",
address = "Budapest, Hungary",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.docinsights-1.21/",
pages = "287--302",
ISBN = "979-8-89176-477-4",
abstract = "Latent context compression replaces a document with a short sequence of learned vectors, but whether this beats simply keeping raw tokens under the same decoder-context budget is rarely tested. We run that test on held-out document QA, matching budgets per example against raw-token truncation and selector-specific readers trained on first-$N$, LLMLingua-2, and question-conditioned BM25 contexts. Our minimal compressor consists of a text encoder, learned pooling, and direct projection. It beats every tested common-cap retention baseline at nominal pooling factors $cr{=}8$ and $cr{=}16$, reaching answer perplexity 3.464 versus 5.349 for the strongest baseline at $cr{=}8$. The advantage persists in first-sentence free generation, and a manual check of answers is directionally consistent. Adding a post-pooling transformer hurts with twelve layers and with one layer. The twelve-layer result holds in the discovery run, two replications, and a second decoder family. Removing the post-pooling transformer also makes the compressor smaller and faster. Learned pooling followed by direct projection provides the strongest quality and efficiency among the tested compressor designs."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="cherepanov-burlova-2026-occupy">
<titleInfo>
<title>What Should Occupy the Context Budget? Latent Compression versus Budget-Matched Truncation in Document QA</title>
</titleInfo>
<name type="personal">
<namePart type="given">Artem</namePart>
<namePart type="family">Cherepanov</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Albina</namePart>
<namePart type="family">Burlova</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>The Proceedings of Workshop on Document Intelligence and Understanding</title>
</titleInfo>
<name type="personal">
<namePart type="given">Santosh</namePart>
<namePart type="family">T.Y.S.S.</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Amit</namePart>
<namePart type="family">Agarwal</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Vivek</namePart>
<namePart type="family">Gupta</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yova</namePart>
<namePart type="family">Kementchedjhieva</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Aparna</namePart>
<namePart type="family">Garimella</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yufang</namePart>
<namePart type="family">Hou</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">979-8-89176-477-4</identifier>
</relatedItem>
<abstract>Latent context compression replaces a document with a short sequence of learned vectors, but whether this beats simply keeping raw tokens under the same decoder-context budget is rarely tested. We run that test on held-out document QA, matching budgets per example against raw-token truncation and selector-specific readers trained on first-N, LLMLingua-2, and question-conditioned BM25 contexts. Our minimal compressor consists of a text encoder, learned pooling, and direct projection. It beats every tested common-cap retention baseline at nominal pooling factors cr=8 and cr=16, reaching answer perplexity 3.464 versus 5.349 for the strongest baseline at cr=8. The advantage persists in first-sentence free generation, and a manual check of answers is directionally consistent. Adding a post-pooling transformer hurts with twelve layers and with one layer. The twelve-layer result holds in the discovery run, two replications, and a second decoder family. Removing the post-pooling transformer also makes the compressor smaller and faster. Learned pooling followed by direct projection provides the strongest quality and efficiency among the tested compressor designs.</abstract>
<identifier type="citekey">cherepanov-burlova-2026-occupy</identifier>
<location>
<url>https://aclanthology.org/2026.docinsights-1.21/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>287</start>
<end>302</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T What Should Occupy the Context Budget? Latent Compression versus Budget-Matched Truncation in Document QA
%A Cherepanov, Artem
%A Burlova, Albina
%Y T.Y.S.S., Santosh
%Y Agarwal, Amit
%Y Gupta, Vivek
%Y Kementchedjhieva, Yova
%Y Garimella, Aparna
%Y Hou, Yufang
%S The Proceedings of Workshop on Document Intelligence and Understanding
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-477-4
%F cherepanov-burlova-2026-occupy
%X Latent context compression replaces a document with a short sequence of learned vectors, but whether this beats simply keeping raw tokens under the same decoder-context budget is rarely tested. We run that test on held-out document QA, matching budgets per example against raw-token truncation and selector-specific readers trained on first-N, LLMLingua-2, and question-conditioned BM25 contexts. Our minimal compressor consists of a text encoder, learned pooling, and direct projection. It beats every tested common-cap retention baseline at nominal pooling factors cr=8 and cr=16, reaching answer perplexity 3.464 versus 5.349 for the strongest baseline at cr=8. The advantage persists in first-sentence free generation, and a manual check of answers is directionally consistent. Adding a post-pooling transformer hurts with twelve layers and with one layer. The twelve-layer result holds in the discovery run, two replications, and a second decoder family. Removing the post-pooling transformer also makes the compressor smaller and faster. Learned pooling followed by direct projection provides the strongest quality and efficiency among the tested compressor designs.
%U https://aclanthology.org/2026.docinsights-1.21/
%P 287-302
Markdown (Informal)
[What Should Occupy the Context Budget? Latent Compression versus Budget-Matched Truncation in Document QA](https://aclanthology.org/2026.docinsights-1.21/) (Cherepanov & Burlova, DocInsights 2026)
ACL