Saved in:
Bibliographic Details
Main Authors: Jin, Shilong, Wang, Lanjun, Zhang, Zhuosheng
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2605.16883
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914572162564096
author Jin, Shilong
Wang, Lanjun
Zhang, Zhuosheng
author_facet Jin, Shilong
Wang, Lanjun
Zhang, Zhuosheng
contents Autonomous Graphical User Interface (GUI) agents often struggle with multi-step tasks due to constrained context windows and static policies that fail to adapt to dynamic environments. To address these limitations, this work proposes the Self-Evolving GUI Agent (SE-GA), a novel framework that integrates hierarchical memory structures with an iterative self-improvement mechanism. At the core of our approach is Test-Time Memory Extension (TTME), which facilitates long-term planning by dynamically retrieving episodic, semantic, and experiential memories to provide salient contexts during inference. To ensure continuous learning, we introduce Memory-Augmented Self-Evolution (MASE), which is a training pipeline that adopts the data collected by TTME to stabilize and enhance the agent's foundational policy. Extensive evaluations across both offline and online benchmarks demonstrate SE-GA achieves state-of-the-art performance, reaching success rates of 89.0\% on ScreenSpot and 75.8\% on the challenging AndroidControl-High dataset. Furthermore, significant improvements on the AndroidWorld benchmark highlight the superior generalization to dynamic environments. Open source code: https://github.com/jinshilong-dev/SE-GA
format Preprint
id arxiv_https___arxiv_org_abs_2605_16883
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle SE-GA: Memory-Augmented Self-Evolution for GUI Agents
Jin, Shilong
Wang, Lanjun
Zhang, Zhuosheng
Machine Learning
Autonomous Graphical User Interface (GUI) agents often struggle with multi-step tasks due to constrained context windows and static policies that fail to adapt to dynamic environments. To address these limitations, this work proposes the Self-Evolving GUI Agent (SE-GA), a novel framework that integrates hierarchical memory structures with an iterative self-improvement mechanism. At the core of our approach is Test-Time Memory Extension (TTME), which facilitates long-term planning by dynamically retrieving episodic, semantic, and experiential memories to provide salient contexts during inference. To ensure continuous learning, we introduce Memory-Augmented Self-Evolution (MASE), which is a training pipeline that adopts the data collected by TTME to stabilize and enhance the agent's foundational policy. Extensive evaluations across both offline and online benchmarks demonstrate SE-GA achieves state-of-the-art performance, reaching success rates of 89.0\% on ScreenSpot and 75.8\% on the challenging AndroidControl-High dataset. Furthermore, significant improvements on the AndroidWorld benchmark highlight the superior generalization to dynamic environments. Open source code: https://github.com/jinshilong-dev/SE-GA
title SE-GA: Memory-Augmented Self-Evolution for GUI Agents
topic Machine Learning
url https://arxiv.org/abs/2605.16883