List of Figures

2.1 Depiction of a thing 15 2.2 Diagram of self-evidencing 16 2.3 Increasing model evidence through action 17 2.4 The agent as the model; inference of hidden causes 19 2.5 Generative model and generative process 20 2.6 A thing with hidden causes; closing the action-perception loop 24 2.7 Conditional independence; deeply hidden causes 27 2.8 The free energy bound on surprise 31 2.9 Heuristic for variational inference 33 3.1 Inference to the best explanation in self-evidencing and the free energy princi ple 53 4.1 Depictions of Markov blankets 72 4.2 Bayesian beliefs about states and state transitions 75 4.3 Precision control within Markov blankets 90 5.1 Self-evidencing as a method 115 7.1 Self-models 151 7.2 Self-models as nested Markov blankets 163 8.1 Active things and strange things 182 9.1 Beliefs about detachment in agents of volatility 226