GUI-CC: Benchmarking Contextual Consistency of GUI World Models as Agent Environments

GUI-CC：评估 GUI 世界模型作为智能体环境的上下文一致性

Abstract: GUI world models are increasingly evaluated as one-step next-screen predictors, yet their intended use is often as multi-step environments for GUI agents. This mismatch leaves a key requirement under-tested: generated states must remain contextually consistent when they are repeatedly reused for future interaction.

摘要： GUI 世界模型越来越多地被评估为单步“下一屏幕”预测器，但其预期的用途往往是作为 GUI 智能体的多步交互环境。这种错位导致一个关键需求未得到充分测试：生成的界面状态在被反复用于后续交互时，必须保持上下文的一致性。

We introduce GUI-CC, a benchmark that evaluates contextual consistency of GUI world models as agent environments rather than isolated next-screen predictors. GUI-CC contains two complementary tracks: an offline reference-action track that rolls models along real mobile GUI trajectories, and an online agent-loop track that lets fixed probing agents interact with model-generated UIs.

我们推出了 GUI-CC，这是一个旨在评估 GUI 世界模型作为智能体环境（而非孤立的下一屏幕预测器）的上下文一致性的基准测试。GUI-CC 包含两个互补的轨道：一个是离线参考动作轨道，通过真实的移动端 GUI 轨迹来运行模型；另一个是在线智能体循环轨道，允许固定的探测智能体与模型生成的 UI 进行交互。

We construct 500 offline trajectory tasks from GUIOdyssey and 200 emulator-verified online tasks across 30 mobile apps. GUI-CC evaluates transition fidelity, transition plausibility, contextual consistency, and task progress.

我们从 GUIOdyssey 中构建了 500 个离线轨迹任务，并跨 30 个移动应用程序构建了 200 个经模拟器验证的在线任务。GUI-CC 对转换保真度、转换合理性、上下文一致性以及任务进度进行评估。

Experiments show that plausible single-step generation does not guarantee reliable environment simulation: current models often produce usable-looking screens while failing to preserve task-relevant context or support executable multi-step rollouts.

实验表明，合理的单步生成并不能保证可靠的环境模拟：当前模型往往能生成看起来可用的屏幕，却无法保持与任务相关的上下文，也无法支持可执行的多步推演。