cs.RONov 18, 2026

Learning to Interact in Multi-Agent Self-Play for Controllable Autonomous Driving Scenario Generation

Authors: Beyazit Yalcinkaya, Ethan Pronovost, Maximilian Naumann, Marcell Vazquez-Chanlatte

Organizations: Zoox, Foster City, USA

Abstract

Large-scale, multi-agent self-play trains robust driving policies effectively in simulation, but controlling these models to generate specific, safety-critical edge cases remains a key challenge. Existing goal-conditioned self-play frameworks rely primarily on static map-location goals and fail to reliably produce complex agent interactions such as aggressive tailgating or adversarial cut-offs. These agent interactions are necessary for rigorous autonomous vehicle (AV) testing. To address this, we present an extended multi-agent self-play framework that introduces agent-relative dynamic goals. By conditioning policies on this extended goal representation, our framework enables targeted, controllable scenario synthesis. We demonstrate that our approach successfully generates interactive driving behaviors that static-goal baselines cannot simulate, providing a flexible tool for long-tail scenario generation in AV testing.

Figures & tables

Explore similar work

CardsList
  1. Heterogeneous Self-Play for Realistic Highway Traffic Simulation

    Mar 31, 2026Jinkai Qiu, Alessandro Saviolo, Chaojie Wang +2Autonomous Driving SimulationAutonomous Driving Scenario Generation

  2. EvoDrive: Pareto Evolution for Safety-Critical Autonomous Driving via Self-Improving LLM Agents

    Jun 2, 2026Tong Nie, Yuewen Mei, Yihong Tang +4Multi-Objective Evolutionary OptimizationAutonomous Driving Safety Evaluation

  3. Composing Driving Worlds through Disentangled Control for Adversarial Scenario Generation

    Mar 13, 2026Yifan Zhan, Zhengqing Chen, Qingjie Wang +7Compositional Video GenerationAutonomous Driving Simulation