@inproceedings{titiya-etal-2026-captr,
title = "{CAPTR}: Caption-based Context Pruning for Multimodal Tabular Reasoning",
author = {Titiya, Prasham and
Engl{\"a}nder, Leon and
Najpande, Gaurav and
Kumar, Tampu Ravi and
Gurevych, Iryna and
Gupta, Vivek},
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.826/",
pages = "19189--19214",
ISBN = "979-8-89176-459-0",
abstract = "Multimodal Large Language Models struggle with tabular reasoning on tables containing images as cells owing to their limited memory and the distractions posed by the surrounding irrelevant visual elements. Existing context-reduction methods reduce context size but degrade reasoning accuracy: text-only pruning neglects visual cells, token-level pruning discards valuable information, and retrieval cannot understand table structure. We propose CAPTR: Caption-based Context Pruning for Tabular Reasoning, a two-stage framework that uses image captions as a lightweight textual proxy to prune non-pertinent rows and columns and then restores images for reasoning. CAPTR can reduce the context length by 80{\%} while achieving up to 8{\%} accuracy improvement against competitive baselines. By analyzing the error cases, we find that entity confusion contributes to the majority of the remaining errors. CAPTR is the first multimodal context-reduction method to improve performance on multimodal tabular reasoning."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="titiya-etal-2026-captr">
<titleInfo>
<title>CAPTR: Caption-based Context Pruning for Multimodal Tabular Reasoning</title>
</titleInfo>
<name type="personal">
<namePart type="given">Prasham</namePart>
<namePart type="family">Titiya</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Leon</namePart>
<namePart type="family">Engländer</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Gaurav</namePart>
<namePart type="family">Najpande</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Tampu</namePart>
<namePart type="given">Ravi</namePart>
<namePart type="family">Kumar</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Iryna</namePart>
<namePart type="family">Gurevych</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Vivek</namePart>
<namePart type="family">Gupta</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>Multimodal Large Language Models struggle with tabular reasoning on tables containing images as cells owing to their limited memory and the distractions posed by the surrounding irrelevant visual elements. Existing context-reduction methods reduce context size but degrade reasoning accuracy: text-only pruning neglects visual cells, token-level pruning discards valuable information, and retrieval cannot understand table structure. We propose CAPTR: Caption-based Context Pruning for Tabular Reasoning, a two-stage framework that uses image captions as a lightweight textual proxy to prune non-pertinent rows and columns and then restores images for reasoning. CAPTR can reduce the context length by 80% while achieving up to 8% accuracy improvement against competitive baselines. By analyzing the error cases, we find that entity confusion contributes to the majority of the remaining errors. CAPTR is the first multimodal context-reduction method to improve performance on multimodal tabular reasoning.</abstract>
<identifier type="citekey">titiya-etal-2026-captr</identifier>
<location>
<url>https://aclanthology.org/2026.emnlp-main.826/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>19189</start>
<end>19214</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T CAPTR: Caption-based Context Pruning for Multimodal Tabular Reasoning
%A Titiya, Prasham
%A Engländer, Leon
%A Najpande, Gaurav
%A Kumar, Tampu Ravi
%A Gurevych, Iryna
%A Gupta, Vivek
%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 titiya-etal-2026-captr
%X Multimodal Large Language Models struggle with tabular reasoning on tables containing images as cells owing to their limited memory and the distractions posed by the surrounding irrelevant visual elements. Existing context-reduction methods reduce context size but degrade reasoning accuracy: text-only pruning neglects visual cells, token-level pruning discards valuable information, and retrieval cannot understand table structure. We propose CAPTR: Caption-based Context Pruning for Tabular Reasoning, a two-stage framework that uses image captions as a lightweight textual proxy to prune non-pertinent rows and columns and then restores images for reasoning. CAPTR can reduce the context length by 80% while achieving up to 8% accuracy improvement against competitive baselines. By analyzing the error cases, we find that entity confusion contributes to the majority of the remaining errors. CAPTR is the first multimodal context-reduction method to improve performance on multimodal tabular reasoning.
%U https://aclanthology.org/2026.emnlp-main.826/
%P 19189-19214
Markdown (Informal)
[CAPTR: Caption-based Context Pruning for Multimodal Tabular Reasoning](https://aclanthology.org/2026.emnlp-main.826/) (Titiya et al., EMNLP 2026)
ACL
- Prasham Titiya, Leon Engländer, Gaurav Najpande, Tampu Ravi Kumar, Iryna Gurevych, and Vivek Gupta. 2026. CAPTR: Caption-based Context Pruning for Multimodal Tabular Reasoning. In Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing, pages 19189–19214, Budapest, Hungary. Association for Computational Linguistics.