GEAR
Generalist Embodied Agent Research
Building Generally Capable Agents in Many Worlds, Virtual and Real
Welcome
Welcome to the NVIDIA GEAR group led by Dr. Jim Fan and Prof. Yuke Zhu. At GEAR, our mission is to build foundation models for embodied agents in virtual and physical worlds. Our research agenda encompasses:
Multimodal Foundation Models
LLMs for planning and reasoning, vision-language models, and world models trained on Internet-scale data
General-Purpose Robots
Robotic models and systems for locomotion and dexterous manipulation in complex environments
Foundation Agents
Large action models that autonomously explore and bootstrap capabilities across games and simulations
Simulation & Synthetic Data
Simulation infrastructure and synthetic data pipelines for large-scale learning
