Scaling Computer-Use Grounding via User Interface Decomposition and Synthesis
Tianbao Xie, Jiaqi Deng, Xiaochuan Li, Junlin Yang, Haoyuan Wu, Jixuan Chen, Wenjing Hu, Xinyuan Wang, Yuhui Xu, Zekun Wang, Yiheng Xu, Junli Wang, Doyen Sahoo, Tao Yu, Caiming Xiong

TL;DR
This paper introduces a comprehensive benchmark and a large dataset for improving GUI grounding, enabling better natural language understanding and manipulation of graphical interfaces, which enhances computer use agents' capabilities.
Contribution
The paper presents OSWorld-G, a detailed benchmark, and Jedi, the largest computer use grounding dataset, advancing the ability to map natural language to complex GUI actions.
Findings
Models trained on Jedi outperform existing approaches on multiple benchmarks.
Grounding improvements significantly enhance agentic performance on complex tasks.
Combining data for different interface elements enables generalization to new interfaces.
Abstract
Graphical user interface (GUI) grounding, the ability to map natural language instructions to specific actions on graphical user interfaces, remains a critical bottleneck in computer use agent development. Current benchmarks oversimplify grounding tasks as short referring expressions, failing to capture the complexity of real-world interactions that require software commonsense, layout understanding, and fine-grained manipulation capabilities. To address these limitations, we introduce OSWorld-G, a comprehensive benchmark comprising 564 finely annotated samples across diverse task types including text matching, element recognition, layout understanding, and precise manipulation. Additionally, we synthesize and release the largest computer use grounding dataset Jedi, which contains 4 million examples through multi-perspective decoupling of tasks. Our multi-scale models trained on Jedi…
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Code & Models
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsTopic Modeling · Multimodal Machine Learning Applications · Natural Language Processing Techniques
