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

Visual HayStack

Benchmarks

Task NameDataset NameSOTA ResultTrend
Long-context Visual Question AnsweringVisual HayStack Long Window (test)
Accuracy (50 images)68.17
11
Long-context Visual Question AnsweringVisual HayStack Short Window (test)
Accuracy (1 Image)98.2
11
Showing 2 of 2 rows