For questions related to the reinforcement learning "dyna" architecture.

For more info, have a look e.g. at https://www.cs.cmu.edu/afs/cs/project/jair/pub/volume4/kaelbling96a-html/node29.html.