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

VLMsAreBlind

Benchmarks

Task NameDataset NameSOTA ResultTrend
Perception & RecognitionVLMsAreBlind
Pass@198.6
7
Visual logicVlmsAreBlind
Top-1 Accuracy66.79
7
Visual ReasoningVLMsAreBlind
VLMsAreBlind Score61.9
6
Multimodal ReasoningVLMsAreBlind
Score (%)52.32
4
Visual PerceptionVlmsAreBlind
Accuracy76.4
2
Showing 5 of 5 rows