GUI Navigation on Web-Single (test)
0.81Type EMGPT-4o-mini
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| GPT-4o-mini2024.06 | 0.81 | 0.623 | 0.57 | |
| MiniCPM-GUIGUIAct=true2024.06 | 0.891 | 0.524 | 0.552 | |
| Fuyu-GUI2024.06 | 0.901 | 0.671 | 0.635 | |
| Qwen-GUIGUIAct=true, GUIChat=true2024.06 | 0.909 | 0.694 | 0.667 | |
| MiniCPM-GUIGUIAct=true, High Resolution=true2024.06 | 0.918 | 0.729 | 0.692 | |
| MiniCPM-GUIGUIAct=true, GUIChat=true2024.06 | 0.918 | 0.749 | 0.706 | |
| Qwen-GUIGUIAct=true2024.06 | 0.919 | 0.656 | 0.65 | |
| MiniCPM-GUIGUIAct=true, Larger LM=true2024.06 | 0.921 | 0.816 | 0.77 | |
| GPT-4o-minitested samples=100 per dataset2024.06 | 81 | 62.3 | 57 | |
| MiniCPM-GUItraining data=GUIAct2024.06 | 89.1 | 52.4 | 55.2 | |
| Fuyu-GUI2024.06 | 90.1 | 67.1 | 63.5 | |
| Qwen-GUI2024.06 | 90.9 | 69.4 | 66.7 | |
| MiniCPM-GUItraining data=GUIAct, configuration=+ High Resolution2024.06 | 91.8 | 72.9 | 69.2 | |
| MiniCPM-GUItraining data=GUIAct, configuration=+ GUIChat2024.06 | 91.8 | 74.9 | 70.6 |