Training Robotic Systems
Sergey and Daniel discuss the surprising findings that training a robotic system can overcome distributional shift issues at test time, even without complicated plans. They explore the effectiveness of surgical perturbations and the potential for large-scale reinforcement learning in designing robotic systems capable of generalizing to various tasks.In this clip
From this podcast
The Gradient
Sergey Levine on Robot Learning & Offline RL
Related Questions
How is machine learning used in robots as discussed in the episode Sergey Levine on Robot Learning & Offline RL and the clip Training Robotic Systems?
How is machine learning used in robots as discussed in the episode Sergey Levine on Robot Learning & Offline RL and the clip Training Robotic Systems from Ep. 37: Sergey Levine on How Deep Learning Will Unleash a Robotics Revolution and the clip Robotics and Learning?
How is machine learning used in robots as discussed in the episode Ep. 37: Sergey Levine on How Deep Learning Will Unleash a Robotics Revolution and the clip Robotics and Learning?