2077AIDataFoundation/VeriGUI

VeriGUI 是一个大规模、人工标注的数据集，旨在促进能够执行复杂、长期任务的自主 GUI 代理的开发和评估。与现有 GUI 数据集专注于短期交互不同，VeriGUI 强调长期复杂性和子任务级可验证性，以更好地反映现实世界中的人机交互场景。VeriGUI is a large-scale, human-annotated dataset designed to facilitate the development and evaluation of autonomous GUI agents capable of performing complex, long-horizon tasks in realistic computer environments. Unlike existing GUI datasets that focus on short-term interactions, VeriGUI emphasizes long-chain complexity and subtask-level verifiability to better reflect real-world human-computer interaction scenarios.