Continual GUI Agents

TL;DR

This paper introduces Continual GUI Agents, a new framework for training GUI agents to adapt to changing environments over time using reinforcement fine-tuning with novel rewards that stabilize learning.

Contribution

It presents GUI-AiF, a reinforcement fine-tuning framework with novel rewards for continual learning in GUI environments, addressing the challenge of shifting UI distributions.

Findings

GUI-AiF outperforms existing methods in experiments.

The framework effectively stabilizes learning amidst domain shifts.

Reinforcement fine-tuning enhances adaptability of GUI agents.

Abstract

As digital environments (data distribution) are in flux, with new GUI data arriving over time-introducing new domains or resolutions-agents trained on static environments deteriorate in performance. In this work, we introduce Continual GUI Agents, a new task that requires GUI agents to perform continual learning under shifted domains and resolutions. We find existing methods fail to maintain stable grounding as GUI distributions shift over time, due to the diversity of UI interaction points and regions in fluxing scenarios. To address this, we introduce GUI-Anchoring in Flux (GUI-AiF), a new reinforcement fine-tuning framework that stabilizes continual learning through two novel rewards: Anchoring Point Reward in Flux (APR-iF) and Anchoring Region Reward in Flux (ARR-iF). These rewards guide the agents to align with shifting interaction points and regions, mitigating the tendency of existing reward strategies to over-adapt to static grounding cues (e.g., fixed coordinates or element scales). Extensive experiments show GUI-AiF surpasses state-of-the-art baselines. Our work establishes the first continual learning framework for GUI agents, revealing the untapped potential of reinforcement fine-tuning for continual GUI Agents.