Do Vision-Language Models See Visualizations Like Humans? Alignment in Chart Categorization | Synapse