Improving LLM spatial reasoning via mobility data

RSS_LABS explores how integrating mobility data into language models enhances their semantic understanding of physical locations and geographic context. The approach links linguistic place-names with movement patterns to ground spatial reasoning.