Jonathan Roberts, Timo Lüddecke, Rehan Sheikh +2
Multimodal large language models (MLLMs) have shown remarkable capabilities across a broad range of tasks but their knowledge and abilities in the geographic and geospatial domains…