{"P": [[[0.0579265376123676, 0.20516758601444435, 0.7369058763731879], [0.7853527806153365, 0.12261166715335005, 0.09203555223131324], [0.2825213785593158, 0.0667012417397371, 0.6507773797009471]], [[0.017155107335849407, 0.11674934705452987, 0.8660955456096207], [0.1288902582047755, 0.6215727295968744, 0.24953701219835012], [0.45953844199848315, 0.20504263949685578, 0.33541891850466105]]], "R": [[0.6962159966701554, 0.2927207490124871], [0.0014900835088361708, 0.9734602747664127], [0.29840122301687566, 0.3139860020343368]], "gamma": 0.99, "states": 3, "actions": 2, "note": "random 3-state, 2-action MDP (state-action-next-state transition probabilities P[a][s][s'], mean rewards R[s][a])"}