1、Develop multimodal models, integrating visual, auditory, sensor, and action data. 2、Research perception-reasoning-action integrated large models to enable robots to autonomously explore and learn. 3、Validate task execution and interaction abilities in real robots or simulated environments.

1、Familiar with deep learning, reinforcement learning, robotics, and multimodal modeling. 2、Experience with robotics or embodied intelligence (ROS, MuJoCo, Isaac Gym, Habitat preferred). 3、Passionate about exploring how AI can “move out of the virtual world and into reality.”