Mobile GUI Automation on AMEX
77.57Type Success RateGUI-R1-7B + CES
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| GUI-R1-7B + CESMethod=Multi-Agent2025.11 | 77.57 | 61.64 | 48.48 | |
| GUI-R1-7B + GPT-5Method=Multi-Agent2025.11 | 72.8 | 52.15 | 35.8 | |
| GUI-R1-7BMethod=RL2025.11 | 67.26 | 57.12 | 43.69 | |
| GUI-Owl-7BMethod=RL2025.11 | 61.56 | 48.38 | 40.48 | |
| UI-R1-3BMethod=RL2025.11 | 60.23 | 41.78 | 35.81 | |
| Qwen2.5-VL-7BMethod=Zero Shot2025.11 | 59.52 | 48.24 | 35.1 | |
| OS-Atlas-7BMethod=SFT2025.11 | 55.11 | 40.3 | 33.89 |