@inproceedings{jung-wang-2026-beyond,
title = "Beyond Chunking: Efficient Global Pooling for Holistic Long-Document Representation",
author = "Jung, Hoin and
Wang, Xiaoqian",
editor = "Lee, Hung-Yi and
Dev, Sunipa and
Li, Jessy and
Elliott, Desmond",
booktitle = "Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing",
month = oct,
year = "2026",
address = "Budapest, Hungary",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.emnlp-main.2080/",
pages = "47372--47397",
ISBN = "979-8-89176-459-0",
abstract = "Effectively representing long documents is a persistent challenge in natural language processing, as foundational encoders are constrained by limited context windows. Prevailing methods like truncation and independent chunk aggregation can discard content or lose cross-chunk context, hindering downstream performance. To address this problem, we introduce **Spectral Attention Token Pooling (SATPool)**, a novel, encoder-agnostic module that generates a single, fixed-size vector from variable-length sequences of locally encoded tokens. SATPool operates in two stages: it first uses efficient linear attention to model global token interactions after chunk encoding, then employs **Spectral Token Compression (STC)** to compress these globally-aware token representations into a compact, context-aware vector. Across long-document classification, Retrieval-Augmented Generation (RAG), and multimodal RAG, SATPool achieves strong overall improvements and robust top-tier performance across datasets and encoder backbones. The pretrained encoder remains frozen, while the lightweight SATPool module and task-specific heads are trained for each downstream task."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="jung-wang-2026-beyond">
<titleInfo>
<title>Beyond Chunking: Efficient Global Pooling for Holistic Long-Document Representation</title>
</titleInfo>
<name type="personal">
<namePart type="given">Hoin</namePart>
<namePart type="family">Jung</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Xiaoqian</namePart>
<namePart type="family">Wang</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>Effectively representing long documents is a persistent challenge in natural language processing, as foundational encoders are constrained by limited context windows. Prevailing methods like truncation and independent chunk aggregation can discard content or lose cross-chunk context, hindering downstream performance. To address this problem, we introduce **Spectral Attention Token Pooling (SATPool)**, a novel, encoder-agnostic module that generates a single, fixed-size vector from variable-length sequences of locally encoded tokens. SATPool operates in two stages: it first uses efficient linear attention to model global token interactions after chunk encoding, then employs **Spectral Token Compression (STC)** to compress these globally-aware token representations into a compact, context-aware vector. Across long-document classification, Retrieval-Augmented Generation (RAG), and multimodal RAG, SATPool achieves strong overall improvements and robust top-tier performance across datasets and encoder backbones. The pretrained encoder remains frozen, while the lightweight SATPool module and task-specific heads are trained for each downstream task.</abstract>
<identifier type="citekey">jung-wang-2026-beyond</identifier>
<location>
<url>https://aclanthology.org/2026.emnlp-main.2080/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>47372</start>
<end>47397</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Beyond Chunking: Efficient Global Pooling for Holistic Long-Document Representation
%A Jung, Hoin
%A Wang, Xiaoqian
%Y Lee, Hung-Yi
%Y Dev, Sunipa
%Y Li, Jessy
%Y Elliott, Desmond
%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 jung-wang-2026-beyond
%X Effectively representing long documents is a persistent challenge in natural language processing, as foundational encoders are constrained by limited context windows. Prevailing methods like truncation and independent chunk aggregation can discard content or lose cross-chunk context, hindering downstream performance. To address this problem, we introduce **Spectral Attention Token Pooling (SATPool)**, a novel, encoder-agnostic module that generates a single, fixed-size vector from variable-length sequences of locally encoded tokens. SATPool operates in two stages: it first uses efficient linear attention to model global token interactions after chunk encoding, then employs **Spectral Token Compression (STC)** to compress these globally-aware token representations into a compact, context-aware vector. Across long-document classification, Retrieval-Augmented Generation (RAG), and multimodal RAG, SATPool achieves strong overall improvements and robust top-tier performance across datasets and encoder backbones. The pretrained encoder remains frozen, while the lightweight SATPool module and task-specific heads are trained for each downstream task.
%U https://aclanthology.org/2026.emnlp-main.2080/
%P 47372-47397
Markdown (Informal)
[Beyond Chunking: Efficient Global Pooling for Holistic Long-Document Representation](https://aclanthology.org/2026.emnlp-main.2080/) (Jung & Wang, EMNLP 2026)
ACL