@inproceedings{wang-etal-2026-vtc,
title = "{VTC}-{R}1: Vision-Text Compression for Efficient Long-Context Reasoning",
author = "Wang, Yibo and
Jing, Yongcheng and
Liu, Shunyu and
Guan, Hao and
Tu, Rong-Cheng and
Wang, Chengyu and
Huang, Jun and
Tao, Dacheng",
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.1560/",
pages = "35560--35578",
ISBN = "979-8-89176-459-0",
abstract = "Long-context reasoning has significantly empowered large language models to tackle complex tasks, yet it introduces severe efficiency bottlenecks due to the computational complexity. Existing efficient approaches often rely on complex additional training or external models for compression, which limits scalability and discards critical fine-grained information. In this paper, we first propose VTC-R1, a new efficient reasoning paradigm that integrates vision-text compression into iterative reasoning. Instead of processing lengthy textual traces, VTC-R1 renders intermediate reasoning segments into compact images, which are iteratively fed back into vision-language models as ``optical memory''. We construct a training dataset based on OpenR1-Math-220K achieving $3.4\times$ token compression and fine-tune representative VLMs{--}Glyph and Qwen3-VL. Extensive experiments on benchmarks such as MATH500, AIME25, AMC23 and GPQA-D demonstrate that VTC-R1 consistently outperforms standard long-context reasoning. And VTC-R1 improves inference efficiency, achieving 2.7{\texttimes} speedup in end-to-end latency, highlighting its potential as a scalable solution for reasoning-intensive applications."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="wang-etal-2026-vtc">
<titleInfo>
<title>VTC-R1: Vision-Text Compression for Efficient Long-Context Reasoning</title>
</titleInfo>
<name type="personal">
<namePart type="given">Yibo</namePart>
<namePart type="family">Wang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yongcheng</namePart>
<namePart type="family">Jing</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Shunyu</namePart>
<namePart type="family">Liu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Hao</namePart>
<namePart type="family">Guan</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Rong-Cheng</namePart>
<namePart type="family">Tu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Chengyu</namePart>
<namePart type="family">Wang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jun</namePart>
<namePart type="family">Huang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Dacheng</namePart>
<namePart type="family">Tao</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>Long-context reasoning has significantly empowered large language models to tackle complex tasks, yet it introduces severe efficiency bottlenecks due to the computational complexity. Existing efficient approaches often rely on complex additional training or external models for compression, which limits scalability and discards critical fine-grained information. In this paper, we first propose VTC-R1, a new efficient reasoning paradigm that integrates vision-text compression into iterative reasoning. Instead of processing lengthy textual traces, VTC-R1 renders intermediate reasoning segments into compact images, which are iteratively fed back into vision-language models as “optical memory”. We construct a training dataset based on OpenR1-Math-220K achieving 3.4\times token compression and fine-tune representative VLMs–Glyph and Qwen3-VL. Extensive experiments on benchmarks such as MATH500, AIME25, AMC23 and GPQA-D demonstrate that VTC-R1 consistently outperforms standard long-context reasoning. And VTC-R1 improves inference efficiency, achieving 2.7× speedup in end-to-end latency, highlighting its potential as a scalable solution for reasoning-intensive applications.</abstract>
<identifier type="citekey">wang-etal-2026-vtc</identifier>
<location>
<url>https://aclanthology.org/2026.emnlp-main.1560/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>35560</start>
<end>35578</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T VTC-R1: Vision-Text Compression for Efficient Long-Context Reasoning
%A Wang, Yibo
%A Jing, Yongcheng
%A Liu, Shunyu
%A Guan, Hao
%A Tu, Rong-Cheng
%A Wang, Chengyu
%A Huang, Jun
%A Tao, Dacheng
%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 wang-etal-2026-vtc
%X Long-context reasoning has significantly empowered large language models to tackle complex tasks, yet it introduces severe efficiency bottlenecks due to the computational complexity. Existing efficient approaches often rely on complex additional training or external models for compression, which limits scalability and discards critical fine-grained information. In this paper, we first propose VTC-R1, a new efficient reasoning paradigm that integrates vision-text compression into iterative reasoning. Instead of processing lengthy textual traces, VTC-R1 renders intermediate reasoning segments into compact images, which are iteratively fed back into vision-language models as “optical memory”. We construct a training dataset based on OpenR1-Math-220K achieving 3.4\times token compression and fine-tune representative VLMs–Glyph and Qwen3-VL. Extensive experiments on benchmarks such as MATH500, AIME25, AMC23 and GPQA-D demonstrate that VTC-R1 consistently outperforms standard long-context reasoning. And VTC-R1 improves inference efficiency, achieving 2.7× speedup in end-to-end latency, highlighting its potential as a scalable solution for reasoning-intensive applications.
%U https://aclanthology.org/2026.emnlp-main.1560/
%P 35560-35578
Markdown (Informal)
[VTC-R1: Vision-Text Compression for Efficient Long-Context Reasoning](https://aclanthology.org/2026.emnlp-main.1560/) (Wang et al., EMNLP 2026)
ACL
- Yibo Wang, Yongcheng Jing, Shunyu Liu, Hao Guan, Rong-Cheng Tu, Chengyu Wang, Jun Huang, and Dacheng Tao. 2026. VTC-R1: Vision-Text Compression for Efficient Long-Context Reasoning. In Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing, pages 35560–35578, Budapest, Hungary. Association for Computational Linguistics.