[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-81639-en":3,"doc-seo-81639-105":29,"detail-sidebar-cat-0-en-105":90},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":4,"is_deleted":4,"is_public":20,"is_downloadable":20,"audit_status":20,"page_count":21,"language":22,"language_code":23,"site_id":24,"html_lang":23,"table_of_contents":25,"faqs":26,"seo_title":13,"seo_description":14,"update_tm":27,"read_time":28},81639,8796095360427,"Lucas Martin","https://ap-avatar.wpscdn.com/davatar_994ba38a5ba835b3df7d355c54d3ed8d",6,"Technology","Render in the Loop Vector Graphics Generation via Visual Self Feedback","Multimodal Large Language Models (MLLMs) generate Scalable Vector Graphics (SVG) by synthesizing symbolic code, but existing open-loop “blind drawing” methods do not render intermediate results, wasting visual priors from vision encoders. This disconnect makes partial canvas reasoning and occlusion relationships difficult, since they are visually explicit yet textually ambiguous. Render-in-the-Loop reformulates SVG synthesis as stepwise, visual-context-aware generation by rendering intermediate code to a cumulative canvas, using Visual Self-Feedback (VSF) and a Render-and-Verify (RaV) inference filter. The framework outperforms open baselines on MMSVGBench for Text-to-SVG and Image-to-SVG.","Render-in-the-Loop: Vector Graphics Generation via Visual Self-Feedback  \nGuotao Liang 1 , Zhangcheng Wang3 , Juncheng Hu 1 , Haitao Zhou 1 , Ziteng Xue 1 , Jing Zhang 1 , Dong Xu2 , and Qian Yu 1 ⋆  \n1 School of Software, Beihang University, China  \n{liangguotao, hujuncheng, zhouhaitao, zt_xue, zhang_jing, [qianyu}@buaa.edu.cn](qianyu}@buaa.edu.cn)  \n2 Department of Computer Science, The University of Hong Kong, China  \ndongxu@cs.hku.hk  \n3 4Paradigm, China  \n[wzc1@mail.ustc.edu.cn](wzc1@mail.ustc.edu.cn)  \narXiv :2604 .20730v3 [ cs .CV] 10 Jul 2026  \nFig. 1: Render-in-the-Loop Generation. (a) Our Visual Self-Feedback explicitly renders intermediate code into a canvas, feeding it back to provide continuous visual guidance. (b) Traditional open-loop approaches draw“blindly” using only textual history, often struggling with geometric accuracy and visual quality. By closing the loop, our method ensures structurally coherent and high-quality SVG synthesis.  \nAbstract. Multimodal Large Language Models (MLLMs) have shown promising capabilities in generating Scalable Vector Graphics (SVG) via direct code synthesis. However, existing paradigms typically adopt an open-loop “blind drawing” approach, where models generate symbolic code sequences without perceiving intermediate visual outcomes. This methodology severely underutilizes the powerful visual priors embedded in MLLMs’ vision encoders, treating SVG generation as a disjointed textual sequence modeling task rather than an integrated visuo-spatial one.  \n⋆ Corresponding author.  \n2 G. Liang et al.  \nConsequently, models struggle to reason about partial canvas states and implicit occlusion relationships, which are visually explicit but textually ambiguous. To bridge this gap, we propose Render-in-the-Loop, a novel generation paradigm that reformulates SVG synthesis as a stepwise, visual-context-aware process. By rendering intermediate code states into a cumulative canvas, the model explicitly observes the evolving visual context at each step, leveraging on-the-fly feedback to guide subsequent generation. However, we demonstrate that applying this visual loop naively to off-the-shelf models is suboptimal due to their inability to leverage incremental visual-code mappings. To address this, we first utilize fine-grained path decomposition to construct dense multi-step visual trajectories, and then introduce a Visual Self-Feedback (VSF) training strategy to condition the next primitive generation on intermediate visual states. Furthermore, a Render-and-Verify (RaV) inference mechanism is proposed to effectively filter degenerate and redundant primitives. Our framework, instantiated on a multimodal foundation model, outperforms strong open-weight baselines on the standard MMSVGBench. This result highlights the remarkable data efficiency and generalization capability of our Render-in-the-Loop paradigm for both Text-to-SVG and Image-toSVG tasks.  \nKeywords: Scalable Vector Graphics · Multimodal Large Language Models · Visual Self-Feedback  \n1 Introduction  \nScalable Vector Graphics (SVG) [6], characterized by resolution independence, high editability, and compact storage, have become an indispensable graphic format in modern UI/UX design, industrial typography, and front-end development [10, 19, 53, 66] . With the rapid progress of Multimodal Large Language Models (MLLMs) in code generation and visual understanding [1, 2], directly leveraging foundation models for Text-to-SVG and Image-to-SVG generation has recently emerged as a promising research direction [39, 58, 67] .  \nPioneering works such as StarVector [39] and OmniSVG [67] have demonstrated that MLLMs are capable of directly producing SVG XML code. However, existing generation paradigms typically adopt an open-loop “blind drawing”approach [5, 39, 58, 60, 67], where the model generates symbolic code sequences without perceiving the intermediate visual outcomes. This methodology severely underutilizes the powerful visual priors embe","cbCaivhajdKixsZK","https://ap.wps.com/l/cbCaivhajdKixsZK","pdf",2830669,1,29,"English","en",105,"# Introduction\n## SVG and MLLMs for Vector Generation\n## Limitations of Open-Loop Blind Drawing\n## Render-in-the-Loop Approach and Feedback Rendering","[{\"question\":\"What problem does Render-in-the-Loop address in SVG generation?\",\"answer\":\"It addresses the limitation of open-loop “blind drawing,” where models generate SVG code without observing intermediate renderings, leading to poor visual coherence and difficulty reasoning about partial canvas states and occlusion.\"},{\"question\":\"How does Render-in-the-Loop use visual feedback during generation?\",\"answer\":\"It renders intermediate code states onto a cumulative canvas and feeds the evolving visual context back to guide the next step of primitive generation.\"},{\"question\":\"What roles do Visual Self-Feedback (VSF) and Render-and-Verify (RaV) play?\",\"answer\":\"VSF trains conditioning on intermediate visual states to improve multi-step primitive generation, while RaV filters degenerate and redundant primitives during inference.\"}]",1784175049,73,{"code":4,"msg":30,"data":31},"ok",{"site_id":24,"language":23,"slug":32,"title":13,"keywords":33,"description":14,"schema_data":34,"social_meta":85,"head_meta":87,"extra_data":89,"updated_unix":27},"render-in-the-loop-vector-graphics-generation-via-visual-self-feedback","",{"@graph":35,"@context":84},[36,53,67],{"@type":37,"itemListElement":38},"BreadcrumbList",[39,43,47,50],{"item":40,"name":41,"@type":42,"position":20},"https://docshare.wps.com","Home","ListItem",{"item":44,"name":45,"@type":42,"position":46},"https://docshare.wps.com/document/","Document",2,{"item":48,"name":12,"@type":42,"position":49},"https://docshare.wps.com/document/technology/",3,{"item":51,"name":13,"@type":42,"position":52},"https://docshare.wps.com/document/render-in-the-loop-vector-graphics-generation-via-visual-self-feedback/81639/",4,{"url":51,"name":13,"@type":54,"author":55,"headline":13,"publisher":57,"fileFormat":60,"inLanguage":23,"description":14,"dateModified":61,"datePublished":61,"encodingFormat":60,"isAccessibleForFree":62,"interactionStatistic":63},"DigitalDocument",{"name":9,"@type":56},"Person",{"url":40,"name":58,"@type":59},"DocShare","Organization","application/pdf","2026-07-16",true,{"@type":64,"interactionType":65,"userInteractionCount":4},"InteractionCounter",{"@type":66},"ViewAction",{"@type":68,"mainEntity":69},"FAQPage",[70,76,80],{"name":71,"@type":72,"acceptedAnswer":73},"What problem does Render-in-the-Loop address in SVG generation?","Question",{"text":74,"@type":75},"It addresses the limitation of open-loop “blind drawing,” where models generate SVG code without observing intermediate renderings, leading to poor visual coherence and difficulty reasoning about partial canvas states and occlusion.","Answer",{"name":77,"@type":72,"acceptedAnswer":78},"How does Render-in-the-Loop use visual feedback during generation?",{"text":79,"@type":75},"It renders intermediate code states onto a cumulative canvas and feeds the evolving visual context back to guide the next step of primitive generation.",{"name":81,"@type":72,"acceptedAnswer":82},"What roles do Visual Self-Feedback (VSF) and Render-and-Verify (RaV) play?",{"text":83,"@type":75},"VSF trains conditioning on intermediate visual states to improve multi-step primitive generation, while RaV filters degenerate and redundant primitives during inference.","https://schema.org",{"og:url":51,"og:type":86,"og:title":13,"og:site_name":58,"og:description":14},"article",{"robots":88,"canonical":51},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":91},[92,96,100,104,109,112,117,122,127,130,134],{"id":20,"doc_module":4,"doc_module_name":45,"category_name":93,"show_sort_weight":94,"slug":95},"Story & Novel",90,"story-novel",{"id":46,"doc_module":4,"doc_module_name":45,"category_name":97,"show_sort_weight":98,"slug":99},"Literature",80,"literature",{"id":52,"doc_module":4,"doc_module_name":45,"category_name":101,"show_sort_weight":102,"slug":103},"Exam",70,"exam",{"id":105,"doc_module":4,"doc_module_name":45,"category_name":106,"show_sort_weight":107,"slug":108},5,"Comic",60,"comic",{"id":11,"doc_module":4,"doc_module_name":45,"category_name":12,"show_sort_weight":110,"slug":111},50,"technology",{"id":113,"doc_module":4,"doc_module_name":45,"category_name":114,"show_sort_weight":115,"slug":116},7,"Healthcare",40,"healthcare",{"id":118,"doc_module":4,"doc_module_name":45,"category_name":119,"show_sort_weight":120,"slug":121},8,"Research & Report",30,"research-report",{"id":123,"doc_module":4,"doc_module_name":45,"category_name":124,"show_sort_weight":125,"slug":126},9,"Religion & Spirituality",20,"religion-spirituality",{"id":125,"doc_module":4,"doc_module_name":45,"category_name":128,"show_sort_weight":125,"slug":129},"World Cup","world-cup",{"id":131,"doc_module":4,"doc_module_name":45,"category_name":132,"show_sort_weight":131,"slug":133},10,"Lifestyle","lifestyle",{"id":135,"doc_module":4,"doc_module_name":45,"category_name":136,"show_sort_weight":105,"slug":137},19,"General","general"]