A simple under-actuated continuous control Reinforcement Learning domain, inspired by the children's toy 'Katita' (https://kikkerland.com/products/katita-wind-up).
![]()
![]()
Please see our GitHub repository for more information and instructions on how to use this model as part of the DeepMind Control Suite.

The Jitterbug Domain v5.0
An under-actuated continuous control Reinforcement Learning domain