Improve model card: add metadata and documentation

#1
by nielsr HF Staff - opened
Files changed (1) hide show
  1. README.md +18 -3
README.md CHANGED
@@ -1,11 +1,25 @@
1
  ---
2
- license: mit
3
  language:
4
  - en
5
- tags:
6
- - arxiv:2602.16855
 
7
  ---
8
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
9
 
10
  ## Citation
11
 
@@ -18,3 +32,4 @@ If you find this model useful, please cite our paper:
18
  journal={arXiv preprint arXiv:2602.16855},
19
  year={2026}
20
  }
 
 
1
  ---
 
2
  language:
3
  - en
4
+ license: mit
5
+ pipeline_tag: image-text-to-text
6
+ library_name: transformers
7
  ---
8
 
9
+ # GUI-Owl-1.5 (Mobile-Agent-v3.5)
10
+
11
+ GUI-Owl-1.5 is a family of native multi-platform GUI agent foundation models built on Qwen3-VL. It supports automation across desktop, mobile, and browser environments, achieving state-of-the-art results on more than 20 GUI benchmarks.
12
+
13
+ - **Paper:** [Mobile-Agent-v3.5: Multi-platform Fundamental GUI Agents](https://arxiv.org/abs/2602.16855)
14
+ - **Code:** [GitHub - X-PLUG/MobileAgent](https://github.com/X-PLUG/MobileAgent)
15
+ - **Demo:** [ModelScope Online Demo](http://modelscope.cn/studios/MobileAgentTest/computer_use)
16
+
17
+ ## Model Description
18
+
19
+ GUI-Owl-1.5 incorporates several key innovations:
20
+ 1. **Hybrid Data Flywheel:** A data pipeline for UI understanding and trajectory generation based on simulated and cloud-based sandbox environments.
21
+ 2. **Unified Enhancement:** A thought-synthesis pipeline to enhance reasoning capabilities, including Tool/MCP use and memory.
22
+ 3. **Multi-platform Environment RL:** The MRPO algorithm to address multi-platform conflicts and improve training efficiency for long-horizon tasks.
23
 
24
  ## Citation
25
 
 
32
  journal={arXiv preprint arXiv:2602.16855},
33
  year={2026}
34
  }
35
+ ```