@inproceedings{zheng-etal-2026-slideforge,
title = "{S}lide{F}orge: Controllable Editing of Slides as Structured Artifacts",
author = "Zheng, Haozhen and
Wang, Fulin and
Xiong, Tianhu and
Yu, YingJie and
Qian, Shengyi and
Yu, Hanchao and
Schwing, Alex and
Nahrstedt, Klara and
Wu, Mingyuan",
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.1690/",
pages = "38429--38471",
ISBN = "979-8-89176-459-0",
abstract = "Current AI agents compellingly describe slides. However, AI-assisted slide editing requires more than understanding: the output must retain layout, style, component structure, and native editability. Towards, AI-assisted slide editing, existing agents operate on screenshots or weak document representations and often fragment coherent visual units, rasterize editable content, or break layout. In contrast, for controllable slide editing we introduce an agentic framework, SlideForge, which builds a Deck State Graph, an executable slide state that links visual decomposition, native PPTX object structure, and perceptual organization. By recovering human-referable components while retaining fine-grained editable structure, SlideForge supports theme-preserving reconstruction through slide-native operations and rendered-state verification. We further introduce an evaluation paradigm for controllable slide transformation that jointly measures component recovery, preservation, restyling consistency, visual quality, and native editability. Experiments show that SlideForge outperforms direct prompting, screenshot-based agents, and generic code-agent baselines across these dimensions."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="zheng-etal-2026-slideforge">
<titleInfo>
<title>SlideForge: Controllable Editing of Slides as Structured Artifacts</title>
</titleInfo>
<name type="personal">
<namePart type="given">Haozhen</namePart>
<namePart type="family">Zheng</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Fulin</namePart>
<namePart type="family">Wang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Tianhu</namePart>
<namePart type="family">Xiong</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">YingJie</namePart>
<namePart type="family">Yu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Shengyi</namePart>
<namePart type="family">Qian</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Hanchao</namePart>
<namePart type="family">Yu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Alex</namePart>
<namePart type="family">Schwing</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Klara</namePart>
<namePart type="family">Nahrstedt</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Mingyuan</namePart>
<namePart type="family">Wu</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>Current AI agents compellingly describe slides. However, AI-assisted slide editing requires more than understanding: the output must retain layout, style, component structure, and native editability. Towards, AI-assisted slide editing, existing agents operate on screenshots or weak document representations and often fragment coherent visual units, rasterize editable content, or break layout. In contrast, for controllable slide editing we introduce an agentic framework, SlideForge, which builds a Deck State Graph, an executable slide state that links visual decomposition, native PPTX object structure, and perceptual organization. By recovering human-referable components while retaining fine-grained editable structure, SlideForge supports theme-preserving reconstruction through slide-native operations and rendered-state verification. We further introduce an evaluation paradigm for controllable slide transformation that jointly measures component recovery, preservation, restyling consistency, visual quality, and native editability. Experiments show that SlideForge outperforms direct prompting, screenshot-based agents, and generic code-agent baselines across these dimensions.</abstract>
<identifier type="citekey">zheng-etal-2026-slideforge</identifier>
<location>
<url>https://aclanthology.org/2026.emnlp-main.1690/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>38429</start>
<end>38471</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T SlideForge: Controllable Editing of Slides as Structured Artifacts
%A Zheng, Haozhen
%A Wang, Fulin
%A Xiong, Tianhu
%A Yu, YingJie
%A Qian, Shengyi
%A Yu, Hanchao
%A Schwing, Alex
%A Nahrstedt, Klara
%A Wu, Mingyuan
%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 zheng-etal-2026-slideforge
%X Current AI agents compellingly describe slides. However, AI-assisted slide editing requires more than understanding: the output must retain layout, style, component structure, and native editability. Towards, AI-assisted slide editing, existing agents operate on screenshots or weak document representations and often fragment coherent visual units, rasterize editable content, or break layout. In contrast, for controllable slide editing we introduce an agentic framework, SlideForge, which builds a Deck State Graph, an executable slide state that links visual decomposition, native PPTX object structure, and perceptual organization. By recovering human-referable components while retaining fine-grained editable structure, SlideForge supports theme-preserving reconstruction through slide-native operations and rendered-state verification. We further introduce an evaluation paradigm for controllable slide transformation that jointly measures component recovery, preservation, restyling consistency, visual quality, and native editability. Experiments show that SlideForge outperforms direct prompting, screenshot-based agents, and generic code-agent baselines across these dimensions.
%U https://aclanthology.org/2026.emnlp-main.1690/
%P 38429-38471
Markdown (Informal)
[SlideForge: Controllable Editing of Slides as Structured Artifacts](https://aclanthology.org/2026.emnlp-main.1690/) (Zheng et al., EMNLP 2026)
ACL
- Haozhen Zheng, Fulin Wang, Tianhu Xiong, YingJie Yu, Shengyi Qian, Hanchao Yu, Alex Schwing, Klara Nahrstedt, and Mingyuan Wu. 2026. SlideForge: Controllable Editing of Slides as Structured Artifacts. In Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing, pages 38429–38471, Budapest, Hungary. Association for Computational Linguistics.