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

CompA

Benchmarks

Task NameDataset NameSOTA ResultTrend
Compositional ReasoningCompA Attribute sub-task
Text Attribute Accuracy44.28
11
Compositional ReasoningCompA Order sub-task
Text Score0.67
11
Audio ReasoningCompA R (test)
Accuracy98.7
2
Showing 3 of 3 rows