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

VPCT

Benchmarks

Task NameDataset NameSOTA ResultTrend
Visual PuzzlesVPCT
Pass@190
7
Visual ReasoningVPCT Ball Drop Maze and Counting
VPCT Accuracy96
5
VPCTVPCT
Human Score4.56
5
Annotation QualityVPCT Ball Drop and Maze Navigation
VPCT Score3.12
5
Annotation-text AlignmentVPCT Ball Drop Maze Navigation
VPCT Score100
5
Showing 5 of 5 rows