Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

AITW

Benchmarks

Task NameDataset NameSOTA ResultTrend
GUI navigationAITW (test)
Install Success Rate78.86
27
GUI NavigationAITW
Overall Success Rate67.3
19
GUI AutomationAITW Gen
PG19.5
12
Mobile Device ControlAitW WebShopping
SR68.3
6
Mobile Device ControlAitW General
Success Rate86.5
6
Mobile Agent Action ExecutionAITW SeeClick
Action Match Score (AMS)62.14
5
UI Task CompletionAITW
Overall Task Time7.7
5
Mobile GUI action matchingAITW instruction-wise (test)
Overall Error7.7
5
GUI agent task planningAITW
General Step Accuracy36.34
4
GUI Agent NavigationAITW (First 128 task samples)
General Success Rate73.2
2
GUI Agent NavigationAITW First 96 task samples
General Success0.7225
2
Showing 11 of 11 rows