UrbanGround benchmark: LLMs' long-range city navigation drops to 0%

机器之心 · wechat · 2026-09-05

Researchers from SJTU, NUS, Meituan, CUHK and Oxford built UrbanGround, a navigable 3D city sandbox from real Hong Kong geodata, with 810 verified tasks testing 10 multimodal models.

The takeaway: models "forget the road after a few steps" — each step's judgment fails to carry through. Paper, code and a playable app are public.

Original post →

More from Multimodal

Multimodal channel →