How well can you do on the map reading spatial reasoning task?

You'll answer 6 questions (about 5 minutes), then see how you did against leading multimodal large language models.

Part of REMAP (REpresentational MAPping), a benchmark evaluating map-to-view spatial reasoning in multimodal large language models, motivated by developmental research on human spatial cognition.

Anonymous responses may be stored for demo testing; no names are collected.