VisualSphinx-V1 is the largest fully-synthetic open-source dataset providing vision logic puzzles.
			
	
	- 
	
	
	VisualSphinx: Large-Scale Synthetic Vision Logic Puzzles for RLPaper β’ 2505.23977 β’ Published β’ 10
- 
	
	
	VisualSphinx/VisualSphinx-V1-RawViewer β’ Updated β’ 662k β’ 302 β’ 3
- 
	
	
	VisualSphinx/VisualSphinx-V1-RL-20KViewer β’ Updated β’ 20k β’ 154 β’ 2
- 
	
	
	VisualSphinx/VisualSphinx-V1-BenchmarkViewer β’ Updated β’ 935 β’ 23
