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

RWQA

Benchmarks

Task NameDataset NameSOTA ResultTrend
Real-world Question AnsweringRWQA
RWQA Accuracy76.99
62
Visual Question AnsweringRWQA
Accuracy80.8
47
Visual Question AnsweringRWQA 158 (val)
Score80.8
23
Visual GroundingRWQA
Accuracy72.29
22
Multimodal UnderstandingRWQA
RWQA Score60.2
14
Multimodal Multi-choiceRWQA
Accuracy70.5
14
Real-world Spatial UnderstandingRWQA
Top-1 Accuracy67.84
10
General EvaluationRWQA
Score71.8
8
Robustness EvaluationRWQA
Accuracy72.9
6
Real-world Multi-modal Question AnsweringRWQA
Accuracy70.46
4
Showing 10 of 10 rows