Reference models

Reference models#

Reference content sits below the core library: worked environments and one replication of a published agent (R-AIF). It reproduces things outside the library and may change without an API break. The core never imports it.

Reference environments#

MountainCar

Sparse-reward continuous Mountain Car.

tmaze

The classic TMaze as a positional process and its parameters, tensor-identical to pymdp's.

Reference replication#

The R-AIF replication of a published agent has its own page, R-AIF.