Code at: <a href=\"https://github.com/YuxiangChai/UI-KOBE\" rel=\"nofollow\">https://github.com/YuxiangChai/UI-KOBE</a></p>\n","updatedAt":"2026-05-29T02:39:27.104Z","author":{"_id":"6458ce236fa580137af5aa95","avatarUrl":"/avatars/db65a7332e375eb5daad5c1b076b1e3b.svg","fullname":"Yuxiang Chai","name":"Yuxiang007","type":"user","isPro":false,"isHf":false,"isHfAdmin":false,"isMod":false,"followerCount":2,"isUserFollowing":false}},"numEdits":0,"identifiedLanguage":{"language":"en","probability":0.33235982060432434},"editors":["Yuxiang007"],"editorAvatarUrls":["/avatars/db65a7332e375eb5daad5c1b076b1e3b.svg"],"reactions":[],"isReport":false}},{"id":"6a1a41444587a78f8e52a7af","author":{"_id":"63d3e0e8ff1384ce6c5dd17d","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/1674830754237-63d3e0e8ff1384ce6c5dd17d.jpeg","fullname":"Librarian Bot (Bot)","name":"librarian-bot","type":"user","isPro":false,"isHf":false,"isHfAdmin":false,"isMod":false,"followerCount":359,"isUserFollowing":false},"createdAt":"2026-05-30T01:45:40.000Z","type":"comment","data":{"edited":false,"hidden":false,"latest":{"raw":"This is an automated message from the [Librarian Bot](https://huggingface.co/librarian-bots). I found the following papers similar to this paper. \n\nThe following papers were recommended by the Semantic Scholar API \n\n* [MobileExplorer: Accelerating On-Device Inference for Mobile GUI Agents via Online Exploration](https://huggingface.co/papers/2605.26546) (2026)\n* [DocOS: Towards Proactive Document-Guided Actions in GUI Agents](https://huggingface.co/papers/2605.18048) (2026)\n* [WebXSkill: Skill Learning for Autonomous Web Agents](https://huggingface.co/papers/2604.13318) (2026)\n* [Executable Agentic Memory for GUI Agent](https://huggingface.co/papers/2605.12294) (2026)\n* [MMSkills: Towards Multimodal Skills for General Visual Agents](https://huggingface.co/papers/2605.13527) (2026)\n* [Towards Scalable Lightweight GUI Agents via Multi-role Orchestration](https://huggingface.co/papers/2604.13488) (2026)\n* [ClawGUI: A Unified Framework for Training, Evaluating, and Deploying GUI Agents](https://huggingface.co/papers/2604.11784) (2026)\n\n\n Please give a thumbs up to this comment if you found it helpful!\n\n If you want recommendations for any Paper on Hugging Face checkout [this](https://huggingface.co/spaces/librarian-bots/recommend_similar_papers) Space\n\n You can directly ask Librarian Bot for paper recommendations by tagging it in a comment: `@librarian-bot recommend`","html":"<p>This is an automated message from the <a href=\"https://huggingface.co/librarian-bots\">Librarian Bot</a>. I found the following papers similar to this paper. </p>\n<p>The following papers were recommended by the Semantic Scholar API </p>\n<ul>\n<li><a href=\"https://huggingface.co/papers/2605.26546\">MobileExplorer: Accelerating On-Device Inference for Mobile GUI Agents via Online Exploration</a> (2026)</li>\n<li><a href=\"https://huggingface.co/papers/2605.18048\">DocOS: Towards Proactive Document-Guided Actions in GUI Agents</a> (2026)</li>\n<li><a href=\"https://huggingface.co/papers/2604.13318\">WebXSkill: Skill Learning for Autonomous Web Agents</a> (2026)</li>\n<li><a href=\"https://huggingface.co/papers/2605.12294\">Executable Agentic Memory for GUI Agent</a> (2026)</li>\n<li><a href=\"https://huggingface.co/papers/2605.13527\">MMSkills: Towards Multimodal Skills for General Visual Agents</a> (2026)</li>\n<li><a href=\"https://huggingface.co/papers/2604.13488\">Towards Scalable Lightweight GUI Agents via Multi-role Orchestration</a> (2026)</li>\n<li><a href=\"https://huggingface.co/papers/2604.11784\">ClawGUI: A Unified Framework for Training, Evaluating, and Deploying GUI Agents</a> (2026)</li>\n</ul>\n<p> Please give a thumbs up to this comment if you found it helpful!</p>\n<p> If you want recommendations for any Paper on Hugging Face checkout <a href=\"https://huggingface.co/spaces/librarian-bots/recommend_similar_papers\">this</a> Space</p>\n<p> You can directly ask Librarian Bot for paper recommendations by tagging it in a comment: <code><span class=\"SVELTE_PARTIAL_HYDRATER contents\" data-target=\"UserMention\" data-props=\"{&quot;user&quot;:&quot;librarian-bot&quot;}\"><span class=\"inline-block\"><span class=\"contents\"><a href=\"/librarian-bot\">@<span class=\"underline\">librarian-bot</span></a></span> </span></span> recommend</code></p>\n","updatedAt":"2026-05-30T01:45:40.497Z","author":{"_id":"63d3e0e8ff1384ce6c5dd17d","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/1674830754237-63d3e0e8ff1384ce6c5dd17d.jpeg","fullname":"Librarian Bot (Bot)","name":"librarian-bot","type":"user","isPro":false,"isHf":false,"isHfAdmin":false,"isMod":false,"followerCount":359,"isUserFollowing":false}},"numEdits":0,"identifiedLanguage":{"language":"en","probability":0.7140752673149109},"editors":["librarian-bot"],"editorAvatarUrls":["https://cdn-avatars.huggingface.co/v1/production/uploads/1674830754237-63d3e0e8ff1384ce6c5dd17d.jpeg"],"reactions":[],"isReport":false}}],"primaryEmailConfirmed":false,"paper":{"id":"2605.29534","authors":[{"_id":"6a18fbd056b4bb14ec65cebb","name":"Yuxiang Chai","hidden":false},{"_id":"6a18fbd056b4bb14ec65cebc","user":{"_id":"666aa99cd1652853e4f9a8b9","avatarUrl":"/avatars/7cd5a0c34b5ccb8eff5a353d88d15a93.svg","isPro":false,"fullname":"HanXiao","user":"HanXiao1999","type":"user","name":"HanXiao1999"},"name":"Han Xiao","status":"claimed_verified","statusLastChangedAt":"2026-05-29T08:51:05.080Z","hidden":false},{"_id":"6a18fbd056b4bb14ec65cebd","name":"Xinyu Fu","hidden":false},{"_id":"6a18fbd056b4bb14ec65cebe","name":"Jinpeng Chen","hidden":false},{"_id":"6a18fbd056b4bb14ec65cebf","name":"Rui Liu","hidden":false},{"_id":"6a18fbd056b4bb14ec65cec0","name":"Hongsheng Li","hidden":false}],"publishedAt":"2026-05-28T00:00:00.000Z","submittedOnDailyAt":"2026-05-29T00:00:00.000Z","title":"UI-KOBE: Knowledge-Oriented Behavior Exploration for Lightweight Graph-Guided GUI Agents","submittedOnDailyBy":{"_id":"6458ce236fa580137af5aa95","avatarUrl":"/avatars/db65a7332e375eb5daad5c1b076b1e3b.svg","isPro":false,"fullname":"Yuxiang Chai","user":"Yuxiang007","type":"user","name":"Yuxiang007"},"summary":"Recent advances in mobile GUI agents have shown strong potential for automating mobile tasks, but most effective systems still depend on large vision-language models for screenshot understanding and long-horizon planning. Small GUI agents that can be deployed directly on mobile devices are more attractive for practical use, offering lower inference cost and better protection of sensitive on-device information. However, due to limited model capacity, such lightweight agents remain unreliable when planning and executing GUI tasks end-to-end from screenshots alone. We propose Knowledge-Oriented Behavior Exploration (UI-KOBE), a framework that improves lightweight mobile GUI agents with reusable app-specific graph knowledge. UI-KOBE first autonomously explores a mobile application and constructs an app knowledge graph, where nodes represent distinct UI states and edges represent executable transitions. At runtime, a lightweight GUI agent uses the graph as external guidance: given a user task and the current screenshot, it identifies the current graph node and selects among self-loop actions, neighboring transitions, task completion, or fallback free actions associated with that node. By supporting runtime decisions with app-specific graph guidance, UI-KOBE reduces the burden of end-to-end GUI planning and helps lightweight models perform mobile GUI tasks more effectively, offering a practical step toward efficient, interpretable, and privacy-conscious on-device GUI agents.","upvotes":7,"discussionId":"6a18fbd056b4bb14ec65cec1","ai_summary":"UI-KOBE framework enhances lightweight mobile GUI agents by incorporating reusable app-specific graph knowledge to improve task planning and execution efficiency.","ai_keywords":["mobile GUI agents","vision-language models","end-to-end GUI planning","app knowledge graph","UI states","executable transitions","runtime decisions","self-loop actions","neighboring transitions","task completion","fallback free actions"]},"canReadDatabase":false,"canManagePapers":false,"canSubmit":false,"hasHfLevelAccess":false,"upvoted":false,"upvoters":[{"_id":"6458ce236fa580137af5aa95","avatarUrl":"/avatars/db65a7332e375eb5daad5c1b076b1e3b.svg","isPro":false,"fullname":"Yuxiang Chai","user":"Yuxiang007","type":"user"},{"_id":"666aa99cd1652853e4f9a8b9","avatarUrl":"/avatars/7cd5a0c34b5ccb8eff5a353d88d15a93.svg","isPro":false,"fullname":"HanXiao","user":"HanXiao1999","type":"user"},{"_id":"630acc9b1d1650ec2f935cc4","avatarUrl":"/avatars/b22fc3d53d9c381236a2760679e49523.svg","isPro":false,"fullname":"Yang","user":"Michael1","type":"user"},{"_id":"64d761b98ebc40443831f82a","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/64d761b98ebc40443831f82a/DHBOtOstiFp2-lDY6b9gb.png","isPro":false,"fullname":"Guangyi Liu","user":"lgy0404","type":"user"},{"_id":"6838123d69b4f4d859bf8f2d","avatarUrl":"/avatars/3a6fc20fa0ce72b009bcec615d0e17a7.svg","isPro":false,"fullname":"Weikang Shi","user":"shiwk24","type":"user"},{"_id":"65a088f4300957620ba45c70","avatarUrl":"/avatars/56ed45e10d3455531979f30881b2d3f9.svg","isPro":false,"fullname":"pengxiang zhao","user":"Pengxiangzhao","type":"user"},{"_id":"660b7374050d4aecdf0f6288","avatarUrl":"/avatars/d313c1aad2179ee718beae695f16d5be.svg","isPro":false,"fullname":"zhidao","user":"bulubulukua","type":"user"}],"acceptLanguages":["en"],"dailyPaperRank":0,"markdownContentUrl":"https://huggingface.co/buckets/huggingchat/papers-content/resolve/2605/2605.29534.md"}">

Get this paper in your agent:

Don't have the latest CLI?

Cite arxiv.org/abs/2605.29534 in a model README.md to link it from this page.

Cite arxiv.org/abs/2605.29534 in a dataset README.md to link it from this page.

Cite arxiv.org/abs/2605.29534 in a Space README.md to link it from this page.

Add this paper to a collection to link it from this page.