2027 Summer Intern, PhD, Research, AV Planning
Core
Design and benchmark systems where LLM meta-agents generate and refine reward functions for autonomous driving planning policies.
Role type
PhD Research Intern (Autonomous Driving Planning)
Builds
Automated reward discovery loops and planner recipes for the Waymo Driver.
Domain
Autonomous driving, Reinforcement Learning, Machine Learning
Deliverable
production ML models
Required skills
Reinforcement Learning (RL/RLFT), Deep Learning policy training, Python, PyTorch or JAX, Autonomous vehicle planning, Trajectory optimization
Preferred skills
LLM-guided code generation, Automated reward discovery, Meta-RL, Peer-reviewed publications at top AI/robotics conferences
Technologies
PyTorch, JAX, LLMs, VLMs
Responsibilities
Design and implement systems for LLM meta-agents to generate reward functions; Diagnose policy failure modes and reward exploitation; Collaborate on transferring validated reward formulations to planner recipes; Document research outcomes and prepare findings for publication.
