Web Browsing Action Prediction on Mind2Web (Cross-Website)
84.8Operation F1OmniParser
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| OmniParserHTML free=true, image=true, variant=LS + ID2024.08 | 84.8 | 41 | 36.5 | |
| Qwen-VLHTML free=true, image=true2024.08 | 83.5 | 13.2 | 9.2 | |
| OmniParserHTML free=true, image=true, variant=LS + GD2024.08 | 83.2 | 41.5 | 36.1 | |
| SeeClickHTML free=true, image=true2024.08 | 80.6 | 21.4 | 16.4 | |
| GPT-4V+textual choiceHTML free=false, image=true2024.08 | 67.8 | 38 | 32.4 | |
| MindActHTML free=false, image=false2024.08 | 65.2 | 42 | 38.9 | |
| GPT-4HTML free=false, image=false2024.08 | 51.1 | 35.8 | 30.1 | |
| GPT-3.5-TurboHTML free=false, image=false2024.08 | 48.8 | 19.3 | 16.2 | |
| MindAct (gen)HTML free=false, image=false2024.08 | 44.7 | 13.9 | 11 | |
| CogAgentHTML free=true, image=true2024.08 | 42.2 | 18.4 | 13.4 | |
| GPT-4V+somHTML free=false, image=true2024.08 | 32.7 | — | 23.7 |