Table 2.

Variables and quantities specific to the writing example of active inference (see main text for details)

Variable	Description
$\alpha \left(t\right)\in {\Re }^6\subset x$	Hidden attractor states: A vector of hidden  states that specify the current location  owards which the agent expects its arm to  be pulled.
$x_i\left(t\right)\in \Re \subset x$ $x_i^{\prime }\left(t\right)\in \Re \subset x$	Hidden effector states: Hidden states that  specify the angular position and velocity of  the i-th joint in a two-jointed arm.
${\ell }_1\left(x_1\right)\in {\Re }^2$ ${\ell }_2(x_1,x_2)\in {\Re }^2$	Joint locations: Locations of the end of the  two arm parts in Cartesian space. These are  functions of the angular positions of the  joints.
${\ell }^{\ast }\left(\alpha \left(t\right)\right)\in {\Re }^2$	Attracting location: The location towards  which the arm is drawn. This is specified by  the hidden attractor states.
$\varphi (x,\alpha )\in {\Re }^2$	Newtonian force: This is the angular force on  the joints exerted by the attracting location.
$A\in {\Re }^{6\times 6}\subset \theta$	Attractor parameters: A matrix of parameters  that govern the (sequential Lotka–Volterra)  dynamics of the hidden attractor states.
$L\in {\Re }^{2\times 6}\subset \theta$	Cartesian parameters: A matrix of  parameters that specify the attracting  locations associated with each hidden  attractor state.