Gengchen Mai

Stanford University, University of California, Santa Barbara

Papers

Total Citations

H-Index

About

Gengchen Mai is a leading researcher at the intersection of artificial intelligence, geographic information science (GIScience), and spatial reasoning. His work centers on developing novel AI architectures—including spatially explicit neural networks and hyperbolic embedding models—to enable machines to perform complex qualitative spatial and temporal reasoning (QSR/QTR). This foundational research addresses a long-standing challenge in AI: moving beyond symbolic reasoning to create systems that can understand and navigate the physical world as humans do. His highly cited paper, "Reasoning over higher-order qualitative spatial relations via spatially explicit neural networks" (2022, 23 citations), is a cornerstone contribution, demonstrating how neural networks can learn and reason about spatial relationships for applications in wayfinding, robotics, and question answering. Mai has also extended these ideas with "HyperQuaternionE" (2022, 11 citations), a hyperbolic embedding model for spatiotemporal reasoning. Demonstrating the broad applicability of his work, he is pioneering the use of Artificial General Intelligence (AGI) for transformative applications, most notably in agriculture ("AGI for Agriculture," 2023, 8 citations), aiming to revolutionize precision farming and resource management. His research is shaping the future of spatially-aware AI.