GUI-explorer: Autonomous Exploration and Mining of Transition-aware Knowledge for GUI Agent

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Xie, Bin, Shao, Rui, Chen, Gongwei, Zhou, Kaiwen, Li, Yinchuan, Liu, Jie, Zhang, Min, Nie, Liqiang
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866909620310638592
author Xie, Bin
Shao, Rui
Chen, Gongwei
Zhou, Kaiwen
Li, Yinchuan
Liu, Jie
Zhang, Min
Nie, Liqiang
author_facet Xie, Bin
Shao, Rui
Chen, Gongwei
Zhou, Kaiwen
Li, Yinchuan
Liu, Jie
Zhang, Min
Nie, Liqiang
contents GUI automation faces critical challenges in dynamic environments. MLLMs suffer from two key issues: misinterpreting UI components and outdated knowledge. Traditional fine-tuning methods are costly for app-specific knowledge updates. We propose GUI-explorer, a training-free GUI agent that incorporates two fundamental mechanisms: (1) Autonomous Exploration of Function-aware Trajectory. To comprehensively cover all application functionalities, we design a Function-aware Task Goal Generator that automatically constructs exploration goals by analyzing GUI structural information (e.g., screenshots and activity hierarchies). This enables systematic exploration to collect diverse trajectories. (2) Unsupervised Mining of Transition-aware Knowledge. To establish precise screen-operation logic, we develop a Transition-aware Knowledge Extractor that extracts effective screen-operation logic through unsupervised analysis the state transition of structured interaction triples (observation, action, outcome). This eliminates the need for human involvement in knowledge extraction. With a task success rate of 53.7% on SPA-Bench and 47.4% on AndroidWorld, GUI-explorer shows significant improvements over SOTA agents. It requires no parameter updates for new apps. GUI-explorer is open-sourced and publicly available at https://github.com/JiuTian-VL/GUI-explorer.
format Preprint
id arxiv_https___arxiv_org_abs_2505_16827
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle GUI-explorer: Autonomous Exploration and Mining of Transition-aware Knowledge for GUI Agent
Xie, Bin
Shao, Rui
Chen, Gongwei
Zhou, Kaiwen
Li, Yinchuan
Liu, Jie
Zhang, Min
Nie, Liqiang
Artificial Intelligence
GUI automation faces critical challenges in dynamic environments. MLLMs suffer from two key issues: misinterpreting UI components and outdated knowledge. Traditional fine-tuning methods are costly for app-specific knowledge updates. We propose GUI-explorer, a training-free GUI agent that incorporates two fundamental mechanisms: (1) Autonomous Exploration of Function-aware Trajectory. To comprehensively cover all application functionalities, we design a Function-aware Task Goal Generator that automatically constructs exploration goals by analyzing GUI structural information (e.g., screenshots and activity hierarchies). This enables systematic exploration to collect diverse trajectories. (2) Unsupervised Mining of Transition-aware Knowledge. To establish precise screen-operation logic, we develop a Transition-aware Knowledge Extractor that extracts effective screen-operation logic through unsupervised analysis the state transition of structured interaction triples (observation, action, outcome). This eliminates the need for human involvement in knowledge extraction. With a task success rate of 53.7% on SPA-Bench and 47.4% on AndroidWorld, GUI-explorer shows significant improvements over SOTA agents. It requires no parameter updates for new apps. GUI-explorer is open-sourced and publicly available at https://github.com/JiuTian-VL/GUI-explorer.
title GUI-explorer: Autonomous Exploration and Mining of Transition-aware Knowledge for GUI Agent
topic Artificial Intelligence
url https://arxiv.org/abs/2505.16827