Finalist
Rank #4
Back to HUD Frontier / RSI RL Environments Hackathon Gallery
Nomos
by Nomos
The end of the traffic light.
Demo Video
About This Project
By 2040 we expect almost every car to drive itself. So why are we still sitting in traffic?
Nomos trained an entire city of self-driving cars to cooperate with multi-agent reinforcement learning on real San Francisco roads. We cut out traffic lights and stop signs. We have hundreds of cars learning to flow together.
Tracks
ML Research
Gaming & Worldsims
Most Viral Winner
Built With
Claude
Deep Learning
HTML/CSS
Jax
Multi-Agent RL
Newtonian Physics
Non-Stationary RL
Python
Reinforcement Learning
Spatial Intelligence
Vercel
Repository
Submitted June 21, 2026 at 12:14 AM