新闻
JarvisGUI: Towards Cross-Device GUI Agents with Dynamic Task Composition
arXiv cs.AI · 发布于 · 阅读约3分钟
30秒读懂
- 发生了什么
- JarvisGUI benchmark evaluates GUI agents on cross-device workflows spanning Android, Windows, and Ubuntu with dynamic task composition.
- 为何重要
- Matters for engineers building or testing GUI automation agents that must handle real-world multi-platform workflows beyond single-device tasks.
- 注意
- Benchmark reveals state-of-the-art open-source agents struggle with state transfer, cross-platform reasoning, and long-horizon dependencies in practice.
- agent
- eval
- benchmark
这条新闻背后的模式
每个模式都讲清楚技术如何运作、何时值得投入，以及在哪里会失效。
The Agent Architect
每周一个模式、一个权衡、一个生产事故案例。为构建智能体系统的人准备的每周简报。
每周一封邮件,一键退订。您的地址仅用于发送简报。