[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-82814-en":3,"doc-seo-82814-105":29,"detail-sidebar-cat-0-en-105":91},{"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":20,"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},82814,4398048950312,"Violet","https://ap-avatar.wpscdn.com/avatar/400002538284de19e3c?_k=1778320343897328908",8,"Research & Report","UI-MOPD: Multi-Platform On-Policy Distillation for Continual GUI Agent Learning","Recent advances in multimodal foundation models and agent systems have shifted GUI agents from single-platform execution to cross-platform interaction, yet building multiplatform GUI agents remains difficult. Cross-platform trajectories are limited and data coverage is narrow, while different platforms use distinct interaction conventions, causing behavioral mixing, capability degradation, and catastrophic forgetting. UI-MOPD introduces multi-teacher on-policy distillation with platform-conditioned teacher routing, enabling adaptation to new platforms while retaining existing skills, validated on OSWorld and MobileWorld.","arXiv :2607 .04425v 1 [ cs .CL] 5 Jul 2026  \nUI-MOPD: Multi-Platform On-Policy Distillation for Continual GUI Agent Learning  \nNiu Lian* 1 ,3 , Alan Chen* 4 , Zhehao Yu3 , Chengzhen Duan2 , Fazhan Liu2 , Hui Liu2 , Pei Fu2 , Jian Luan2 , Yaowei Wang3 ,5 , Shu-Tao Xia 1 ,5 , Jinpeng Wang★3  \n1Tsinghua Shenzhen International Graduate School, Tsinghua University, 2Xiaomi  \n3 Harbin Institute of Technology, Shenzhen, 4 Zhejiang University, 5 Peng Cheng Laboratory  \n* Equal contribution. ★ Corresponding author.  \n[220110904@stu.hit.edu.cn](220110904@stu.hit.edu.cn) (Niu Lian), [wangjp26@gmail.com](wangjp26@gmail.com) (Jinpeng Wang)  \n§ GitHub  Hugging Face § Homepage  \nAbstract  \nRecent advances in multimodal foundation models and agent systems have driven GUI agents from single-platform task execution toward cross-platform interaction. However, building multiplatform GUI agents remains challenging. On one hand, high-quality and executable cross-platform interaction trajectories are still scarce, and existing data often suffer from limited platform coverage. On the other hand, different platforms exhibit distinct interaction conventions, making joint or continual training prone to behavioral pattern mixing, platform-specific capability degradation, and catastrophic forgetting. To address these challenges, we construct Uni-GUI, a high-quality cross-platform GUI interaction dataset, and propose UI-MOPD, the first method that incorporates multi-teacher on-policy distillation into continual learning for GUI agents. UI-MOPD dynamically selects a platform-specific teacher according to the current environment and transfers platformspecific behavioral priors to a shared policy through platform-conditioned distillation, enabling adaptation to new platforms while preserving capabilities on existing ones. Experiments on OSWorld and MobileWorld show that UI-MOPD achieves task success rates of 38.2% and 12.0%, respectively, demonstrating its effectiveness in balancing cross-platform capability retention and new-platform adaptation.  \n\n| (a) Model Merge |  | Parameter-Space Merge\u003Cbr>\u003Cbr>Desktop Teacher Mobile Teacher Action-Space Conflict Desktop Actions Mobile Actions |\n| --- | --- | --- |\n|  | Desktop Teacher\u003Cbr>\u003Cbr>\u003Cbr>Mobile Teacher |  |\n\n\n| (b) Mixed SFT |  |\n| --- | --- |\n|  | Result: Action Convention Collapse\u003Cbr>Desktop Actions\u003Cbr>Averaged Policy\u003Cbr>Mobile Actions |\n\n(c) MOPD (Ours)  \nDesktop Env Mobile Env  \nDesktop Teacher  \nMobile Teacher  \nFigure 1 Motivation of UI-MOPD. Naively combining desktop and mobile signals, as in model merging or mixed SFT, can mix platform-specific behavioral conventions and produce an averaged policy. UI-MOPD uses platform-conditioned routing and multi-teacher on-policy distillation to integrate platform-specific expertise into a shared GUI agent.  \nContents  \n1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3  \n2 Related Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4  \n2.1 GUI Agent: From Single-plantform to Multi-plantform ...................... 4  \n2.2 Multi-Teacher On-Policy Distillation ................................ 4  \n3 Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4  \n3. 1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5  \n3.2 Multi-Teacher On-Policy Distillation ................................ 5  \n3.3 Platform-Conditioned Teacher Routing ............................... 6  \n3.4 Reward Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7  \n3.5 Training Objective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7  \n4 Experiments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7  \n4. 1 Overview . . . . . . . . . . . . . . . . . . . . . . . . ","cbCaitCcCWdv2U9i","https://ap.wps.com/l/cbCaitCcCWdv2U9i","pdf",3213508,1,25,"English","en",105,"# Contents\n## Introduction\n## Related Work\n## Method\n## Experiments\n## Conclusion\n## Appendix","[{\"question\":\"What problem does UI-MOPD address in continual GUI agent learning?\",\"answer\":\"It addresses the difficulty of training GUI agents across multiple platforms without mixing platform-specific behavior conventions, degrading capabilities, or causing catastrophic forgetting.\"},{\"question\":\"How does UI-MOPD integrate multi-teacher on-policy distillation?\",\"answer\":\"UI-MOPD dynamically selects a platform-specific teacher based on the current environment and transfers platform-specific behavioral priors to a shared policy via platform-conditioned distillation.\"},{\"question\":\"What performance results does UI-MOPD achieve on OSWorld and MobileWorld?\",\"answer\":\"Experiments report task success rates of 38.2% on OSWorld and 12.0% on MobileWorld, showing effective balancing between cross-platform capability retention and adaptation to new platforms.\"}]",1784183142,63,{"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":86,"head_meta":88,"extra_data":90,"updated_unix":27},"ui-mopd-multi-platform-on-policy-distillation-for-continual-gui-agent-learning","",{"@graph":35,"@context":85},[36,53,68],{"@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/research-report/",3,{"item":51,"name":13,"@type":42,"position":52},"https://docshare.wps.com/document/ui-mopd-multi-platform-on-policy-distillation-for-continual-gui-agent-learning/82814/",4,{"url":51,"name":13,"@type":54,"author":55,"headline":13,"publisher":57,"fileFormat":60,"inLanguage":23,"description":14,"dateModified":61,"datePublished":62,"encodingFormat":60,"isAccessibleForFree":63,"interactionStatistic":64},"DigitalDocument",{"name":9,"@type":56},"Person",{"url":40,"name":58,"@type":59},"DocShare","Organization","application/pdf","2026-07-17","2026-07-16",true,{"@type":65,"interactionType":66,"userInteractionCount":20},"InteractionCounter",{"@type":67},"ViewAction",{"@type":69,"mainEntity":70},"FAQPage",[71,77,81],{"name":72,"@type":73,"acceptedAnswer":74},"What problem does UI-MOPD address in continual GUI agent learning?","Question",{"text":75,"@type":76},"It addresses the difficulty of training GUI agents across multiple platforms without mixing platform-specific behavior conventions, degrading capabilities, or causing catastrophic forgetting.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"How does UI-MOPD integrate multi-teacher on-policy distillation?",{"text":80,"@type":76},"UI-MOPD dynamically selects a platform-specific teacher based on the current environment and transfers platform-specific behavioral priors to a shared policy via platform-conditioned distillation.",{"name":82,"@type":73,"acceptedAnswer":83},"What performance results does UI-MOPD achieve on OSWorld and MobileWorld?",{"text":84,"@type":76},"Experiments report task success rates of 38.2% on OSWorld and 12.0% on MobileWorld, showing effective balancing between cross-platform capability retention and adaptation to new platforms.","https://schema.org",{"og:url":51,"og:type":87,"og:title":13,"og:site_name":58,"og:description":14},"article",{"robots":89,"canonical":51},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":92},[93,97,101,105,110,115,120,123,128,131,135],{"id":20,"doc_module":4,"doc_module_name":45,"category_name":94,"show_sort_weight":95,"slug":96},"Story & Novel",90,"story-novel",{"id":46,"doc_module":4,"doc_module_name":45,"category_name":98,"show_sort_weight":99,"slug":100},"Literature",80,"literature",{"id":52,"doc_module":4,"doc_module_name":45,"category_name":102,"show_sort_weight":103,"slug":104},"Exam",70,"exam",{"id":106,"doc_module":4,"doc_module_name":45,"category_name":107,"show_sort_weight":108,"slug":109},5,"Comic",60,"comic",{"id":111,"doc_module":4,"doc_module_name":45,"category_name":112,"show_sort_weight":113,"slug":114},6,"Technology",50,"technology",{"id":116,"doc_module":4,"doc_module_name":45,"category_name":117,"show_sort_weight":118,"slug":119},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":45,"category_name":12,"show_sort_weight":121,"slug":122},30,"research-report",{"id":124,"doc_module":4,"doc_module_name":45,"category_name":125,"show_sort_weight":126,"slug":127},9,"Religion & Spirituality",20,"religion-spirituality",{"id":126,"doc_module":4,"doc_module_name":45,"category_name":129,"show_sort_weight":126,"slug":130},"World Cup","world-cup",{"id":132,"doc_module":4,"doc_module_name":45,"category_name":133,"show_sort_weight":132,"slug":134},10,"Lifestyle","lifestyle",{"id":136,"doc_module":4,"doc_module_name":45,"category_name":137,"show_sort_weight":106,"slug":138},19,"General","general"]