Generalised free energy and active inference

@mjdramstead @bayesianboy @CarcassiFausto @a_tschantz @BerenMillidge @manuelbaltieri @KaiUeltzhoeffer https://t.co/LfJgC3pTZG

@CarcassiFausto From point of view of variational Bayes, the expected free energy term G is a bit of a sleight of hand because it is included as an empirical prior. However, Parr & Friston have rederived its form by considering a more formal 'generalis

@36zimmer @NathanielVirgo @_fernando_rosas @a_tschantz @manuelbaltieri @BerenMillidge The new third wave formulations are currently being explored, and include the ‘generalized free energy’ (https://t.co/EQVTA9zCe3); the ‘free-energy of the expected future

@NathanielVirgo @manuelbaltieri @_fernando_rosas @BerenMillidge "Generalised" free energy goes some way to combining these into a single functional (https://t.co/pOwgOS1xGt), but I think it's a work in progress.