Research Intern - World-Action Foundation Model, Robotics
Core
Conduct research on pretraining world-action foundation models integrating vision, physics, and language modalities to enable robot actions and simulation world generation.
Role type
Research Intern (World-Action Foundation Model, Robotics)
Builds
Pretrained world-action foundation models and simulation environments for autonomous driving and robotic generalist applications.
Domain
Physical AI, Robotics, Autonomous Driving, Computer Vision, Graphics
Deliverable
production ML models | research
Required skills
Machine learning, Computer vision, Graphics, Robotics, Spatial reasoning, Deformable object modeling, Simulation
Preferred skills
Vision-physics modality association, Feed-forward Gaussian splatting, Human data incorporation, Language modality integration
Technologies
Gaussian splatting, World foundation models
Responsibilities
Conduct research on pretraining world-action foundation models with various world modalities; Dive into topics such as vision-physics modality association, feed-forward Gaussian splatting, and spatial reasoning; Work closely with Research Scientists and Engineers on high quality research publications for submission to top-tier conferences
Seniority
Intern
