Drawing inspiration from animal motor control allows effective training of a single locomotion policy for diverse quadruped robots.
