| | --- |
| | license: mit |
| | --- |
| | |
| | # UI-AGILE: Advancing GUI Agents with Effective Reinforcement Learning and Precise Inference-Time Grounding |
| |
|
| | <font size=4><div align='center' > [[📖 Paper](https://arxiv.org/abs/2507.22025)] [[🤗 Checkpoints](https://huggingface.co/KDEGroup/UI-AGILE)] [[🤗 Data](https://huggingface.co/datasets/KDEGroup/UI-AGILE-Data)] [[🤗 Daily Paper](https://huggingface.co/papers/2507.22025)] [[🚀 Github](https://github.com/KDEGroup/UI-AGILE)]</div></font> |
| |
|
| | ## 🔥 Overview |
| |
|
| | UI-AGILE enhances GUI agents through improved training with a Continuous Reward function, Simple Thinking reward, and **Cropping-based Resampling**, and inference with **Decomposed Grounding with Selection**. |
| | |
| |
|
| | [[🤗 UI-AGILE-3B](https://huggingface.co/KDEGroup/UI-AGILE-3B)] |
