Model
153d agoS2Vec Maps Cities With Language Models
Researchers at Google AI unveiled S2Vec, a graph‑based language model that interprets urban layouts as linguistic structures. By translating streets, transit hubs, and zoning patterns into semantic embeddings, the system can predict traffic flows and inform city planning worldwide.
The Signal
The approach shows how AI can read and reason about complex cityscapes, offering a new tool for global urban analytics.