Hugging Face Daily Papers · September 4, 2026 · 3 min read

Editable Visual Design

Mirrored from Hugging Face Daily Papers for archival readability. Support the source by reading on the original site.

Like Read original ↗

Github: <a href=\"https://github.com/yejy53/Editable-Design\" rel=\"nofollow\">https://github.com/yejy53/Editable-Design</a></p>\n","updatedAt":"2026-09-04T02:18:08.176Z","author":{"_id":"6039478ab3ecf716b1a5fd4d","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/6039478ab3ecf716b1a5fd4d/_Thy4E7taiSYBLKxEKJbT.jpeg","fullname":"taesiri","name":"taesiri","type":"user","isPro":true,"isHf":false,"isHfAdmin":false,"isMod":false,"followerCount":370,"isUserFollowing":false}},"numEdits":0,"identifiedLanguage":{"language":"en","probability":0.5837504863739014},"editors":["taesiri"],"editorAvatarUrls":["https://cdn-avatars.huggingface.co/v1/production/uploads/6039478ab3ecf716b1a5fd4d/_Thy4E7taiSYBLKxEKJbT.jpeg"],"reactions":[],"isReport":false}}],"primaryEmailConfirmed":false,"paper":{"id":"2609.04034","authors":[{"_id":"6a9a2a3c8f7c3b7557239475","user":{"_id":"66978ee0b8656f6506b4acb2","avatarUrl":"/avatars/298acb8222e189fce4368985ee5374a1.svg","isPro":false,"fullname":"Junyan Ye","user":"Yejy53","type":"user","name":"Yejy53"},"name":"Junyan Ye","status":"claimed_verified","statusLastChangedAt":"2026-09-04T08:45:04.246Z","hidden":false},{"_id":"6a9a2a3c8f7c3b7557239476","name":"Wei Liu","hidden":false},{"_id":"6a9a2a3c8f7c3b7557239477","user":{"_id":"6349214f8146350b3a4c5cdf","avatarUrl":"/avatars/cfd24caac9a87efb528d0f4c375932bc.svg","isPro":false,"fullname":"Dongzhi Jiang","user":"CaraJ","type":"user","name":"CaraJ"},"name":"Dongzhi Jiang","status":"claimed_verified","statusLastChangedAt":"2026-09-04T08:45:04.251Z","hidden":false},{"_id":"6a9a2a3c8f7c3b7557239478","name":"Zichen Wen","hidden":false},{"_id":"6a9a2a3c8f7c3b7557239479","name":"HaoDong Li","hidden":false},{"_id":"6a9a2a3c8f7c3b755723947a","name":"Zhutao Lv","hidden":false},{"_id":"6a9a2a3c8f7c3b755723947b","name":"Jiaxin Lin","hidden":false},{"_id":"6a9a2a3c8f7c3b755723947c","name":"Jinhua Yu","hidden":false},{"_id":"6a9a2a3c8f7c3b755723947d","user":{"_id":"670ddb69d6ac6394419d88c5","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/no-auth/XxnGNaX3FWug4aiZVjg93.png","isPro":false,"fullname":"Jun He","user":"JunHe0915","type":"user","name":"JunHe0915"},"name":"Jun He","status":"claimed_verified","statusLastChangedAt":"2026-09-04T08:45:04.256Z","hidden":false},{"_id":"6a9a2a3c8f7c3b755723947e","user":{"_id":"6487e158f675b4a7867f45fa","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/6487e158f675b4a7867f45fa/J0sls6zZ682o-SH7iQs7B.jpeg","isPro":false,"fullname":"Zilong Huang","user":"SereinH","type":"user","name":"SereinH"},"name":"Zilong Huang","status":"claimed_verified","statusLastChangedAt":"2026-09-04T08:45:04.261Z","hidden":false},{"_id":"6a9a2a3c8f7c3b755723947f","name":"Rui Chen","hidden":false},{"_id":"6a9a2a3c8f7c3b7557239480","name":"Weijia Li","hidden":false}],"publishedAt":"2026-09-03T00:00:00.000Z","submittedOnDailyAt":"2026-09-04T00:00:00.000Z","title":"Editable Visual Design","submittedOnDailyBy":{"_id":"6039478ab3ecf716b1a5fd4d","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/6039478ab3ecf716b1a5fd4d/_Thy4E7taiSYBLKxEKJbT.jpeg","isPro":true,"fullname":"taesiri","user":"taesiri","type":"user","name":"taesiri"},"summary":"While diffusion base models such as GPT-Image-2 and Nano-Banana exhibit remarkable visual expressiveness, their end-to-end generation inherently yields flattened bitmaps with error-prone text, precluding layer-wise post-editing. Conversely, code-based visual generation via Coding Agents provides precise layout control and decoupled layers, yet remains constrained by a lack of global aesthetic intuition and the difficulty of coding complex visual assets.\n To address this, we propose Editable Visual Design, a new paradigm driven by a Coding Agent. We designate the VLM as the ``creative brain'' for requirement comprehension, task planning, and aesthetic judgment, while utilizing the image generation model as an on-demand ``visual world simulator'' to synthesize standalone visual assets. Operating under an ``imagine first, then act'' closed-loop workflow, the agent generates isolated assets, writes native HTML/CSS, and iteratively refines the design against visual rendering feedback.\n Furthermore, Agent Design Replay faithfully reproduces the creative and reasoning trajectory akin to that of professional human designers. Ultimately, the system delivers editable artifacts with decoupled layers and real text, enabling users to perform intuitive mouse dragging and layout adjustments on a graphical user interface. Validations on posters, infographics, and other scenarios show that this paradigm successfully achieves both refined aesthetics and production-grade editability.","upvotes":22,"discussionId":"6a9a2a3d8f7c3b7557239481","ai_summary":"A coding agent guided by a vision-language model generates editable layered designs by synthesizing isolated visual assets and iteratively refining native HTML/CSS layouts.","ai_keywords":["diffusion base models","VLM","Coding Agent","visual world simulator","imagine first then act","Agent Design Replay","HTML/CSS","editable artifacts"],"ai_summary_model":"thinkingmachines/Inkling-Small","organization":{"_id":"6645f953c39288df638dbdd5","name":"Tencent-Hunyuan","fullname":"Tencent Hunyuan","avatar":"https://cdn-avatars.huggingface.co/v1/production/uploads/62d22496c58f969c152bcefd/woKSjt2wXvBNKussyYPsa.png"}},"canReadDatabase":false,"canManagePapers":false,"canSubmit":false,"hasHfLevelAccess":false,"upvoted":false,"upvoters":[{"_id":"66978ee0b8656f6506b4acb2","avatarUrl":"/avatars/298acb8222e189fce4368985ee5374a1.svg","isPro":false,"fullname":"Junyan Ye","user":"Yejy53","type":"user"},{"_id":"6487e158f675b4a7867f45fa","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/6487e158f675b4a7867f45fa/J0sls6zZ682o-SH7iQs7B.jpeg","isPro":false,"fullname":"Zilong Huang","user":"SereinH","type":"user"},{"_id":"6349214f8146350b3a4c5cdf","avatarUrl":"/avatars/cfd24caac9a87efb528d0f4c375932bc.svg","isPro":false,"fullname":"Dongzhi Jiang","user":"CaraJ","type":"user"},{"_id":"61454a989cd783fec339bdd0","avatarUrl":"/avatars/39cc15c0a70e0d2b1f1ef1c7a98e7db8.svg","isPro":false,"fullname":"Xi Yang","user":"ianyeung","type":"user"},{"_id":"67a62f8e47fc008de27889ea","avatarUrl":"/avatars/0b3cba784863962b1cc10d99de810d59.svg","isPro":false,"fullname":"Wei Liu","user":"willie1206","type":"user"},{"_id":"6877917cb2d7c0a48afb5487","avatarUrl":"/avatars/ef3ea56c925537a6ab745cea1a8b8067.svg","isPro":false,"fullname":"Yuncheng Guo","user":"Yunncheng","type":"user"},{"_id":"65ddea8b2d26e59a5a33330f","avatarUrl":"/avatars/3104ddafd6dda3c05ea9a771dbf2deeb.svg","isPro":false,"fullname":"li haodong","user":"mickyhimself","type":"user"},{"_id":"670dff999b7725cd97b59942","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/noauth/E-wjQxu8ckEcIG_tkb0Ac.jpeg","isPro":false,"fullname":"yi","user":"edged","type":"user"},{"_id":"670ddb69d6ac6394419d88c5","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/no-auth/XxnGNaX3FWug4aiZVjg93.png","isPro":false,"fullname":"Jun He","user":"JunHe0915","type":"user"},{"_id":"660d755f51ed58ee308a9de1","avatarUrl":"/avatars/6af003114889127a3f76397ea92c35c6.svg","isPro":false,"fullname":"Xu","user":"xujinghua","type":"user"},{"_id":"66c598b536a75deef8bb21c0","avatarUrl":"/avatars/b3721202d01d3c947c36f0beae43ea1c.svg","isPro":false,"fullname":"Hengrui Kang","user":"khr0516","type":"user"},{"_id":"6555e4dc598cbe7bfdcde31f","avatarUrl":"/avatars/5b07aee1fd66268905d6c266ef678a34.svg","isPro":false,"fullname":"Thea","user":"TheaFu","type":"user"}],"acceptLanguages":["en"],"dailyPaperRank":0,"organization":{"_id":"6645f953c39288df638dbdd5","name":"Tencent-Hunyuan","fullname":"Tencent Hunyuan","avatar":"https://cdn-avatars.huggingface.co/v1/production/uploads/62d22496c58f969c152bcefd/woKSjt2wXvBNKussyYPsa.png"},"markdownContentUrl":"https://huggingface.co/buckets/huggingchat/papers-content/resolve/2609/2609.04034.md","query":{}}">

Papers

arxiv:2609.04034

Editable Visual Design

Published on Sep 3

· Submitted by

taesiri on Sep 4

Tencent Hunyuan

Upvote

Authors:

Junyan Ye ,

Dongzhi Jiang ,

Jun He ,

Zilong Huang ,

Abstract

A coding agent guided by a vision-language model generates editable layered designs by synthesizing isolated visual assets and iteratively refining native HTML/CSS layouts.

Generated by thinkingmachines/Inkling-Small

While diffusion base models such as GPT-Image-2 and Nano-Banana exhibit remarkable visual expressiveness, their end-to-end generation inherently yields flattened bitmaps with error-prone text, precluding layer-wise post-editing. Conversely, code-based visual generation via Coding Agents provides precise layout control and decoupled layers, yet remains constrained by a lack of global aesthetic intuition and the difficulty of coding complex visual assets. To address this, we propose Editable Visual Design, a new paradigm driven by a Coding Agent. We designate the VLM as the ``creative brain'' for requirement comprehension, task planning, and aesthetic judgment, while utilizing the image generation model as an on-demand ``visual world simulator'' to synthesize standalone visual assets. Operating under an ``imagine first, then act'' closed-loop workflow, the agent generates isolated assets, writes native HTML/CSS, and iteratively refines the design against visual rendering feedback. Furthermore, Agent Design Replay faithfully reproduces the creative and reasoning trajectory akin to that of professional human designers. Ultimately, the system delivers editable artifacts with decoupled layers and real text, enabling users to perform intuitive mouse dragging and layout adjustments on a graphical user interface. Validations on posters, infographics, and other scenarios show that this paradigm successfully achieves both refined aesthetics and production-grade editability.

View arXiv page View PDF Add to collection