Reinforcement Learning and World Model for Autonomous Simulation Intern - 2027
Core
Develop multi-modal world models and integrate them into simulation systems to enable autonomous driving vehicles to interpret, anticipate, and respond in dynamic contexts.
Role type
PhD-level research intern in Reinforcement Learning and World Models
Builds
Multi-modal world simulation models, self-supervised latent dynamics models, and sensor generation models for autonomous driving
Domain
Autonomous driving, Physical AI, Computer Graphics
Deliverable
production ML models
Required skills
Reinforcement Learning (policy gradients, actor-critic, offline RL), Neural Rendering, 4D Scene Representation (NeRF, Gaussian Splatting, occupancy networks), Generative Models (diffusion, flow matching), Large-scale training pipelines, Sim2Real transfer
Preferred skills
Publications in RL or autonomous systems, Open-source contributions
Technologies
NeRF, Gaussian Splatting, Diffusion models, Flow matching, Policy gradients
Responsibilities
Develop and refine multi-modal world models; Train and evaluate self-supervised latent dynamics and sensor generation models; Collaborate with End-to-End Driving Model teams to deploy world-model-based policies; Contribute to system development for continuous learning and simulation adaptation
Seniority
PhD candidate (Research Intern)