MobileGUI-RL: Advancing Mobile GUI Agent through Reinforcement Learning in Online Environment
This addresses scalability and robustness issues for mobile GUI automation, though it is incremental as it builds on existing reinforcement learning methods.
The paper tackled the problem of training vision-based GUI agents in offline environments, which limits scalability and causes overfitting, by introducing MobileGUI-RL, a framework that trains agents online using a curriculum and adapted reinforcement learning, resulting in consistent gains on three benchmarks.
Recently, there has been a surge of vision-based GUI agents designed to automate everyday mobile and web tasks. These agents interpret raw GUI screenshots and autonomously decide where to click, scroll, or type, which bypasses handcrafted rules and app-specific APIs. However, most existing methods trained GUI agent in the offline environment using pre-collected trajectories. This approach limits scalability, causes overfitting to specific UI templates, and leads to brittle policies when faced with unseen environment. We present MobileGUI-RL, a scalable framework that trains GUI agent in online environment. MobileGUI-RL contains two key components. It (i) synthesizes a curriculum of learnable tasks through self-exploration and filtering, and (ii) adapts GRPO to GUI navigation with trajectory-aware advantages and composite rewards that balance task success and execution efficiency. Experiments on three online mobile-agent benchmarks show consistent gains, validating the effectiveness of our approach.