BEAP-Agent: Backtrackable Execution and Adaptive Planning for GUI Agents

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Lu, Ziyu, Weng, Tengjin, Yang, Yiying, Zhao, Yuhang, Huang, Xinxin, Jiang, Wenhao
Format: Preprint
Veröffentlicht: 2026
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866918312972124160
author Lu, Ziyu
Weng, Tengjin
Yang, Yiying
Zhao, Yuhang
Huang, Xinxin
Jiang, Wenhao
author_facet Lu, Ziyu
Weng, Tengjin
Yang, Yiying
Zhao, Yuhang
Huang, Xinxin
Jiang, Wenhao
contents GUI agents are designed to automate repetitive tasks and enhance productivity. However, existing GUI agents struggle to recover once they follow an incorrect exploration path, often leading to task failure. In this work, we model GUI task execution as a DFS process and propose BEAP-Agent, a DFS-based framework that supports long-range, multi-level state backtracking with dynamic task tracking and updating. The framework consists of three collaborative components: Planner, Executor, and Tracker. Together, they enable effective task exploration and execution. BEAP-Agent fills the gap in systematic backtracking mechanisms for GUI agents, offering a systematic solution for long-horizon task exploration. We conducted a systematic evaluation on the OSWorld benchmark, where BEAP-Agent achieved an accuracy of 28.2%, validating the effectiveness of the proposed method.
format Preprint
id arxiv_https___arxiv_org_abs_2601_21352
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle BEAP-Agent: Backtrackable Execution and Adaptive Planning for GUI Agents
Lu, Ziyu
Weng, Tengjin
Yang, Yiying
Zhao, Yuhang
Huang, Xinxin
Jiang, Wenhao
Artificial Intelligence
GUI agents are designed to automate repetitive tasks and enhance productivity. However, existing GUI agents struggle to recover once they follow an incorrect exploration path, often leading to task failure. In this work, we model GUI task execution as a DFS process and propose BEAP-Agent, a DFS-based framework that supports long-range, multi-level state backtracking with dynamic task tracking and updating. The framework consists of three collaborative components: Planner, Executor, and Tracker. Together, they enable effective task exploration and execution. BEAP-Agent fills the gap in systematic backtracking mechanisms for GUI agents, offering a systematic solution for long-horizon task exploration. We conducted a systematic evaluation on the OSWorld benchmark, where BEAP-Agent achieved an accuracy of 28.2%, validating the effectiveness of the proposed method.
title BEAP-Agent: Backtrackable Execution and Adaptive Planning for GUI Agents
topic Artificial Intelligence
url https://arxiv.org/abs/2601.21352