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

GAR-Bench

Benchmarks

Task NameDataset NameSOTA ResultTrend
Visual Question AnsweringGAR-Bench-VQA
Overall VQA Score64.2
17
Localized relational captioningGAR-Bench Cap
Overall Score62.2
15
Showing 2 of 2 rows