Recently, there has been a growing interest in modeling planning with
information constraints. Accordingly, an agent maximizes a regularized expected
utility known as the free energy, where the regularizer is given by the
information divergence from a prior to a posterior policy. While