Learning to Drive in New Cities Without Human Demonstrations

arXiv:2602.15891 · cs.RO, cs.LG, cs.MA · Submitted 2026-02-09 · Read on arXiv

cs.RO, cs.LG, cs.MA

Submitted: 2026-02-09

Updated: 2026-10-22

Comments: Autonomous Driving, Reinforcement Learning, Self-play, Simulation, Transfer Learning, Data-efficient Adaptation. Project page: https://zerlinwang.github.io/NOMAD/

Code: https://github.com/Emerge-Lab/PufferDrive

Project page: https://zerlinwang.github.io/NOMAD

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Terminology

• Adaptation — 1× in this paper · explained in 1 episode
• Reinforcement Learning — 1× in this paper · explained in 4 episodes
• Reward Function — 1× in this paper · explained in 4 episodes

Sources

• Scaling Laws of Motion Forecasting and Planning -- Technical Report
• NuPlan: A closed-loop ML-based planning benchmark for autonomous vehicles
• SPACeR: Self-Play Anchoring with Centralized Reference Models
• Human-compatible driving partners through data-regularized self-play reinforcement learning
• Building reliable sim driving agents by scaling self-play
• Robust Autonomy Emerges from Self-Play
• Is Independent Learning All You Need in the StarCraft Multi-Agent Challenge?
• ORION: A Holistic End-to-End Autonomous Driving Framework by Vision-Language Instructed Action Generation
• GAIA-1: A Generative World Model for Autonomous Driving
• Benchmarking Robustness in Object Detection: Autonomous Driving when Winter is Coming
• Cosmos-Drive-Dreams: Scalable Synthetic Driving Data Generation with World Foundation Models
• GAIA-2: A Controllable Multi-View Generative World Model for Autonomous Driving
• TeraSim-World: Worldwide Safety-Critical Data Synthesis for End-to-End Autonomous Driving
• RoCA: Robust Cross-Domain End-to-End Autonomous Driving
• A Study on Overfitting in Deep Reinforcement Learning

Related papers

• FMT x: An Efficient and Asymptotically Optimal Extension of the Fast Marching Tree for Dynamic Replanning
• MPCFormer: A physics-informed data-driven approach for explainable socially-aware autonomous driving
• RoboLab: A High-Fidelity Simulation Benchmark for Analysis of Task Generalist Policies
• ARC: A Reasoning Recipe for Robot Foundation Models
• HRDexDB: A 4D Dexterous Grasping Dataset Across Human and Multiple Robot Embodiments
• APT: Action Expert Pretraining Improves Instruction Generalization of Vision-Language-Action Policies