Table 1.

Definitions of the tuple Embedded Image underlying active inference.

a sample space Ω or non-empty set from which random fluctuations or outcomes Embedded Image are drawn
external states Embedded Image—hidden states of the world that cause sensory states and depend on action
sensory states Embedded Image—signals that constitute a probabilistic mapping from action and external states
active states Embedded Image—action that depends on sensory and internal states
internal states Embedded Image—representational states that cause action and depend on sensory states
ergodic density Embedded Image—a probability density function over external Embedded Image, sensory Embedded Image, active Embedded Image and internal states Embedded Image for a Markov blanket denoted by m
variational density Embedded Image—an arbitrary probability density function over external states that is parametrized by internal states