[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-84578-en":3,"doc-seo-84578-105":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},84578,8796095360427,"Lucas Martin","https://ap-avatar.wpscdn.com/davatar_994ba38a5ba835b3df7d355c54d3ed8d",8,"Research & Report","A Task-State Representation for Long-Horizon Mobile GUI Agents","Long-horizon mobile GUI agents often use thought-action-observation loops, but the prompt blends persistent task state with transient screen observations. As execution histories grow, this entanglement increases context load, leading to goal drift, progress hallucination, and repeated interactions with stale interfaces. Task-State Representation (TSR) introduces an explicit, training-free decoupling between task state and sensory input. TSR updates via pre- and post-action visual comparisons and maintains a global instruction summary, a dynamic progress tracker, and a transition-aware action verifier. Experiments on four benchmarks show up to a 12-point success-rate improvement.","A Task-State Representation for Long-Horizon Mobile GUI Agents  \nYujie Zheng2∗, Zikang Liu 1∗, Xin Zhao 1†, Ji-Rong Wen 1  \n1 Gaoling School of Artificial Intelligence, Renmin University of China  \n2 School of Software, Beihang University {janeyujie555,jasonlaw8121,[batmanfly}@gmail.com](batmanfly}@gmail.com) , [jrwen@ruc.edu.cn](jrwen@ruc.edu.cn)  \narXiv :2607 .00502v 1 [ cs .CL] 1 Jul 2026  \nAbstract  \nWhile long-horizon mobile GUI agents typically rely on thought-action-observation loops, they struggle to separate persistent task states from transient screen observations. As execution histories grow, this entanglement imposesa severe context burden, causing agents to forget initial requirements, hallucinate progress, or repeatedly interact with stale interfaces. To address this, we introduce Task-State Representation (TSR)—a training-free framework that explicitly decouples task state from sensory input. Acting as a lightweight external wrapper, TSR maintains three structured components: a global instruction summary, a dynamic progress tracker for subgoals, and a transitionaware action verifier. By continuously updating through pre-and post-action visual comparisons, TSR effectively guides the agent’s reasoning without requiring architectural modifications. Experiments across four mobile GUI benchmarks validate TSR’s effectiveness, yielding up to a 12 absolute point increase in success rate on complex cross-application and memoryintensive tasks.  \n1 Introduction  \nAutomating mobile tasks via graphical user interfaces (GUIs) remains a long-standing goal in the development of intelligent agents. Recent multimodal large language models (MLLMs) (Liu et al., 2023) have facilitated prompt-based GUI actors that observe screenshots, reason about the current state, and generate executable actions, such as tapping, typing, or scrolling (Zhang et al., 2025 ; Hong et al., 2024) . The dominant paradigm for handling long-horizon tasks adopts a thought-actionobservation (Yao et al., 2022) loop: at each step, the actor receives the task instruction, a window of recent screenshots, and a history of previous reasoning and actions, and subsequently generates a  \n*Equal contribution.  \n†Corresponding author.  \nnew reasoning trace followed by an action (Zhang et al., 2024 ; Rawles et al., 2025) . This append-all design relies on an implicit assumption—that the actor can reliably maintain awareness of the overall task goal and the cumulative progress from anever-growing raw trajectory.  \nIn practice, this assumption breaks down as the steps increases. We identify three recurring failure modes in long-horizon mobile benchmarks:  \n(1) goal drift, where the actor gradually loses sight of the original task after observing numerous intermediate screens; (2) progress hallucination, where the actor lose access to earlier visual observations and fabricates past states when reasoning about cumulative progress; and (3) stale-screen repetition, where the actor misinterprets a delayed update in the user interface as a failed action and enters a localized loop.  \nThese failures are not solely attributable to limitations in the model’s capacity. Rather, they arise from a structural deficiency in how the input of the actor is organized: the standard prompt conflates two fundamentally different categories of information—persistent task state (the request of the user, the accomplished subgoals, and the remaining steps) and transient observation state (the content displayed on the current screen) . Without an explicit mechanism to separate and maintain the former, the actor must re-derive the task progress from the raw history at every step—a burden that scales linearly with the length of the trajectory.  \nWe propose a task-state representation that addresses this separation. The representation is maintained externally to a fixed GUI actor and is updated at each step by a training-free state updater that compares the pre-action and post-action screenshots. It com","cbCaibwm8bxJqDUn","https://ap.wps.com/l/cbCaibwm8bxJqDUn","pdf",3469587,1,9,"English","en",105,"# Abstract\n# Introduction\n# Method\n## Problem Formulation","[{\"question\":\"What problem does TSR address in long-horizon mobile GUI agents?\",\"answer\":\"TSR targets the entanglement of persistent task state with transient screen observations, which grows into context overload and causes goal drift, progress hallucination, and stale-screen repetition.\"},{\"question\":\"How does TSR separate task state from screen observations?\",\"answer\":\"TSR provides an external, training-free task-state block that is updated by comparing pre-action and post-action screenshots, then injected into the GUI actor’s prompt.\"},{\"question\":\"What components does the TSR task-state block maintain?\",\"answer\":\"TSR maintains three structured views: a global instruction summary, a dynamic progress tracker for subgoals, and a transition-aware action verifier.\"}]",1784196901,23,{"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},"a-task-state-representation-for-long-horizon-mobile-gui-agents","",{"@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/a-task-state-representation-for-long-horizon-mobile-gui-agents/84578/",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 TSR address in long-horizon mobile GUI agents?","Question",{"text":75,"@type":76},"TSR targets the entanglement of persistent task state with transient screen observations, which grows into context overload and causes goal drift, progress hallucination, and stale-screen repetition.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"How does TSR separate task state from screen observations?",{"text":80,"@type":76},"TSR provides an external, training-free task-state block that is updated by comparing pre-action and post-action screenshots, then injected into the GUI actor’s prompt.",{"name":82,"@type":73,"acceptedAnswer":83},"What components does the TSR task-state block maintain?",{"text":84,"@type":76},"TSR maintains three structured views: a global instruction summary, a dynamic progress tracker for subgoals, and a transition-aware action verifier.","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},[93,97,101,105,110,115,120,123,127,130,134],{"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":21,"doc_module":4,"doc_module_name":45,"category_name":124,"show_sort_weight":125,"slug":126},"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":106,"slug":137},19,"General","general"]