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

VisMin

Benchmarks

Task NameDataset NameSOTA ResultTrend
Vision-Language CompositionalityVisMin Hard (test)
Text Score (T)84.47
16
Compositional ReasoningVisMin Easy (test)
Text Score94.88
16
Visual ReasoningVisMin (test)
Text Score (T)91.06
2
Showing 3 of 3 rows