BacktrackAgent: Enhancing GUI Agent with Error Detection and Backtracking Mechanism

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Wu, Qinzhuo, Gao, Pengzhi, Liu, Wei, Luan, Jian
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866913860034756608
author Wu, Qinzhuo
Gao, Pengzhi
Liu, Wei
Luan, Jian
author_facet Wu, Qinzhuo
Gao, Pengzhi
Liu, Wei
Luan, Jian
contents Graphical User Interface (GUI) agents have gained substantial attention due to their impressive capabilities to complete tasks through multiple interactions within GUI environments. However, existing agents primarily focus on enhancing the accuracy of individual actions and often lack effective mechanisms for detecting and recovering from errors. To address these shortcomings, we propose the BacktrackAgent, a robust framework that incorporates a backtracking mechanism to improve task completion efficiency. BacktrackAgent includes verifier, judger, and reflector components as modules for error detection and recovery, while also applying judgment rewards to further enhance the agent's performance. Additionally, we develop a training dataset specifically designed for the backtracking mechanism, which considers the outcome pages after action executions. Experimental results show that BacktrackAgent has achieved performance improvements in both task success rate and step accuracy on Mobile3M and Auto-UI benchmarks. Our data and code will be released upon acceptance.
format Preprint
id arxiv_https___arxiv_org_abs_2505_20660
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle BacktrackAgent: Enhancing GUI Agent with Error Detection and Backtracking Mechanism
Wu, Qinzhuo
Gao, Pengzhi
Liu, Wei
Luan, Jian
Computation and Language
Artificial Intelligence
Graphical User Interface (GUI) agents have gained substantial attention due to their impressive capabilities to complete tasks through multiple interactions within GUI environments. However, existing agents primarily focus on enhancing the accuracy of individual actions and often lack effective mechanisms for detecting and recovering from errors. To address these shortcomings, we propose the BacktrackAgent, a robust framework that incorporates a backtracking mechanism to improve task completion efficiency. BacktrackAgent includes verifier, judger, and reflector components as modules for error detection and recovery, while also applying judgment rewards to further enhance the agent's performance. Additionally, we develop a training dataset specifically designed for the backtracking mechanism, which considers the outcome pages after action executions. Experimental results show that BacktrackAgent has achieved performance improvements in both task success rate and step accuracy on Mobile3M and Auto-UI benchmarks. Our data and code will be released upon acceptance.
title BacktrackAgent: Enhancing GUI Agent with Error Detection and Backtracking Mechanism
topic Computation and Language
Artificial Intelligence
url https://arxiv.org/abs/2505.20660