@inproceedings{zhang-etal-2026-thinking-visual,
title = "Thinking with Visual Grounding",
author = "Zhang, Junkai and
Deng, Yihe and
Chang, Kai-Wei and
Wang, Wei",
editor = "Lee, Hung-Yi and
Dev, Sunipa and
Li, Jessy and
Elliott, Desmond",
booktitle = "Findings of the {A}ssociation for {C}omputational {L}inguistics: {EMNLP} 2026",
month = oct,
year = "2026",
address = "Budapest, Hungary",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.findings-emnlp.1418/",
pages = "30224--30237",
ISBN = "979-8-89176-462-0",
abstract = "Visual thinking should not only sound right; it should show its evidence. While recent vision-language models (VLMs) can produce natural language thinking traces, these traces often leave the supporting image regions implicit, making them hard to verify and weakly supervised. We introduce \textit{visually grounded thinking}, a thinking process in which models interleave natural language thoughts with explicit point or box grounding to the visual evidence used at each step. This makes the model explains its intermediate thoughts in language while grounding key object in the image regions they refer to. To train this behavior, we construct a scalable synthesis pipeline that distills correct visual reasoning traces, extracts the visual objects required by the trace, grounds them with a SAM3-based agent, and derives aligned point and box supervision from the resulting masks. We further propose grounding-aware reinforcement learning, which combines answer rewards with dense grounding rewards that score whether generated object references match the correct image evidence. Across two counting benchmarks and four spatial reasoning benchmarks, adding visually grounded thinking to Gemma3-4B-IT consistently improves performance over the original model and the non-grounded thinking baseline. On spatial reasoning, the visually grounded thinking 4B models match, and in some cases surpass, Gemma3-27B-IT from the same model family. Our analysis shows that point grounding is well suited to counting, while box grounding benefits most from explicit grounding rewards on spatial tasks. Overall, our results show that VLMs think better when their intermediate thoughts are tied to the image regions that make them true."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="zhang-etal-2026-thinking-visual">
<titleInfo>
<title>Thinking with Visual Grounding</title>
</titleInfo>
<name type="personal">
<namePart type="given">Junkai</namePart>
<namePart type="family">Zhang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yihe</namePart>
<namePart type="family">Deng</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Kai-Wei</namePart>
<namePart type="family">Chang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Wei</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>Findings of the Association for Computational Linguistics: EMNLP 2026</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">979-8-89176-462-0</identifier>
</relatedItem>
<abstract>Visual thinking should not only sound right; it should show its evidence. While recent vision-language models (VLMs) can produce natural language thinking traces, these traces often leave the supporting image regions implicit, making them hard to verify and weakly supervised. We introduce visually grounded thinking, a thinking process in which models interleave natural language thoughts with explicit point or box grounding to the visual evidence used at each step. This makes the model explains its intermediate thoughts in language while grounding key object in the image regions they refer to. To train this behavior, we construct a scalable synthesis pipeline that distills correct visual reasoning traces, extracts the visual objects required by the trace, grounds them with a SAM3-based agent, and derives aligned point and box supervision from the resulting masks. We further propose grounding-aware reinforcement learning, which combines answer rewards with dense grounding rewards that score whether generated object references match the correct image evidence. Across two counting benchmarks and four spatial reasoning benchmarks, adding visually grounded thinking to Gemma3-4B-IT consistently improves performance over the original model and the non-grounded thinking baseline. On spatial reasoning, the visually grounded thinking 4B models match, and in some cases surpass, Gemma3-27B-IT from the same model family. Our analysis shows that point grounding is well suited to counting, while box grounding benefits most from explicit grounding rewards on spatial tasks. Overall, our results show that VLMs think better when their intermediate thoughts are tied to the image regions that make them true.</abstract>
<identifier type="citekey">zhang-etal-2026-thinking-visual</identifier>
<location>
<url>https://aclanthology.org/2026.findings-emnlp.1418/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>30224</start>
<end>30237</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Thinking with Visual Grounding
%A Zhang, Junkai
%A Deng, Yihe
%A Chang, Kai-Wei
%A Wang, Wei
%Y Lee, Hung-Yi
%Y Dev, Sunipa
%Y Li, Jessy
%Y Elliott, Desmond
%S Findings of the Association for Computational Linguistics: EMNLP 2026
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-462-0
%F zhang-etal-2026-thinking-visual
%X Visual thinking should not only sound right; it should show its evidence. While recent vision-language models (VLMs) can produce natural language thinking traces, these traces often leave the supporting image regions implicit, making them hard to verify and weakly supervised. We introduce visually grounded thinking, a thinking process in which models interleave natural language thoughts with explicit point or box grounding to the visual evidence used at each step. This makes the model explains its intermediate thoughts in language while grounding key object in the image regions they refer to. To train this behavior, we construct a scalable synthesis pipeline that distills correct visual reasoning traces, extracts the visual objects required by the trace, grounds them with a SAM3-based agent, and derives aligned point and box supervision from the resulting masks. We further propose grounding-aware reinforcement learning, which combines answer rewards with dense grounding rewards that score whether generated object references match the correct image evidence. Across two counting benchmarks and four spatial reasoning benchmarks, adding visually grounded thinking to Gemma3-4B-IT consistently improves performance over the original model and the non-grounded thinking baseline. On spatial reasoning, the visually grounded thinking 4B models match, and in some cases surpass, Gemma3-27B-IT from the same model family. Our analysis shows that point grounding is well suited to counting, while box grounding benefits most from explicit grounding rewards on spatial tasks. Overall, our results show that VLMs think better when their intermediate thoughts are tied to the image regions that make them true.
%U https://aclanthology.org/2026.findings-emnlp.1418/
%P 30224-30237
Markdown (Informal)
[Thinking with Visual Grounding](https://aclanthology.org/2026.findings-emnlp.1418/) (Zhang et al., Findings 2026)
ACL
- Junkai Zhang, Yihe Deng, Kai-Wei Chang, and Wei Wang. 2026. Thinking with Visual Grounding. In Findings of the Association for Computational Linguistics: EMNLP 2026, pages 30224–30237, Budapest, Hungary. Association for Computational Linguistics.