neurologyHigh Demand
Embodied Foundation Models
Train and deploy large-scale multimodal models that understand physical world dynamics for robotic task planning and execution.
View RolesIntelligent agents that learn through physical interaction — sim-to-real transfer, embodied reasoning, and foundation models for the real world.
Train and deploy large-scale multimodal models that understand physical world dynamics for robotic task planning and execution.
View RolesBridge the reality gap by developing domain randomization, system identification, and transfer learning techniques for physical deployment.
View RolesDesign agents that learn manipulation and navigation skills through trial-and-error interaction in both simulated and real environments.
View Roles