# SecAgent: Efficient Mobile GUI Agent with Semantic Context

Source: [SecAgent: Efficient Mobile GUI Agent with Semantic Context](<https://www.sota2.com/research/paper/69afa5c12d603e6de3e43f95>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-04-02T05:00:04.464Z

Authors: Yiping Xie, Song Chen, Jingxuan Xing, Wei Jiang, Zekun Zhu, Yingyao Wang, Pi Bu, Jun Song, Yuning Jiang, Bo Zheng

Published: 2026-03-09T16:04:08.000Z

## Abstract

Mobile Graphical User Interface (GUI) agents powered by multimodal large language models have demonstrated promising capabilities in automating complex smartphone tasks. However, existing approaches face two critical limitations: the scarcity of high-quality multilingual datasets, particularly for non-English ecosystems, and inefficient history representation methods. To address these challenges, we present SecAgent, an efficient mobile GUI agent at 3B scale. We first construct a human-verified Chinese mobile GUI dataset with 18k grounding samples and 121k navigation steps across 44 applications, along with a Chinese navigation benchmark featuring multi-choice action annotations. Building upon this dataset, we propose a semantic context mechanism that distills history screenshots and actions into concise, natural language summaries, significantly reducing computational costs while preserving task-relevant information. Through supervised and reinforcement fine-tuning, SecAgent outperforms similar-scale baselines and achieves performance comparable to 7B-8B models on our and public navigation benchmarks. Our dataset is available at https://huggingface.co/datasets/alibabagroup/CMGUI.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

- [GUI Interaction Navigation on CMGUI](<https://www.sota2.com/research/sota/gui-interaction-navigation-on-cmgui.md>): GUI interaction navigation; CMGUI; Step Accuracy: 96.4.

- [Mobile Navigation on GUIOdyssey updated evaluation set](<https://www.sota2.com/research/sota/mobile-navigation-on-guiodyssey-updated-evaluation-set.md>): Mobile navigation; GUIOdyssey updated evaluation set; Step Accuracy: 74.3.

- [Mobile navigation on AndroidControl](<https://www.sota2.com/research/sota/mobile-navigation-on-androidcontrol.md>): Mobile navigation; AndroidControl; Step Accuracy: 69.5.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2603.08533>)

[PDF](<https://arxiv.org/pdf/2603.08533.pdf>)
