Research Scientist, RL for Autonomous Planning & World Modeling

This role centers on researching and developing cutting-edge reinforcement learning and distillation methods to enable autonomous trajectory planning and robust world modeling.
You will participate in foundation world model post-training, evaluation, and integration into scalable internal RL pipelines.
You will implement and compare RL techniques for autonomous planning, conducting thorough ablations to identify practical, scalable methods.
You will collaborate with engineering and research teams to share recipes, post-training best practices, and contribute to shared infrastructure.
You will work across related topics such as multi-modal learning, Bayesian inference, and robust evaluation to push capabilities under diverse driving conditions.
Applicants should have a PhD or Masters with 3+ years in RL or foundation models and a track record of impactful contributions to the field.