Phi Fluctuates with Surprise: An empirical pre-study for the synthesis of the Free Energy Principle and Integrated Information Theory

Abstract

The Free Energy Principle (FEP) and Integrated Information Theory (IIT) are two ambitious theoretical frameworks, the first aiming to make a general formal description of self-organization and life-like processes, and the second attempting a mathematical theory of conscious experience based on the intrinsic properties of a system. They are each concerned with complementary aspects of the properties of systems, one with life and behavior the other with meaning and experience, so combining them has potentially great scientific value. In this paper, we take a first step towards this synthesis by first partially replicating the results of the evolutionary simulation study by Albantakis et al. (2014) that show a relationship between IIT-measures and fitness in differing complexities of tasks. We then relate FEP-related information theoretic measures to this result, finding that the surprisal of simulated agents’ system states follows the general increase in fitness over evolutionary time, and that it fluctuates together with IIT-based consciousness measures in within-trial time. This suggests that the consciousness measures of IIT indirectly depend on the relation between the agent and the external world, and that they therefore should be related to concepts directly used in the FEP. Lastly, we suggest a future approach for investigating this relationship empirically.

Time consciousness: the missing link in theories of consciousness

“…After a proliferation of alternative theories of consciousness in recent decades, the current trend is toward parsimony through synthesis ( Waade et al 2020 ), standardization ( Graziano et al 2020 ), or direct competition ( Reardon 2019 ). In this vein, Northoff and Lamme (2020) reviewed the convergence between eight theories of consciousness (i.e.…”

Bayesian theories of consciousness: a review in search for a minimal unifying model

“… 14 While IIT has been less represented among the views discussed here than other competing theories, a recent preprint ( Waade et al 2020 ) discusses another Bayesian approach to consciousness currently under development, which is likely to strengthen the position of IIT in this context. …”

Predictive neuronal adaptation as a basis for consciousness.

“…Here we illustrate how the above stages of conscious competence could be recapitulated by the network with our predictive learning rule. We used the neuronal surprise as a proxy measure of consciousness, which is motivated by previous theoretical (Friston, 2018;Waade et al, 2020) and experimental work (Babiloni et al, 2006;Del Cul et al, 2007), which will be discussed in later sections. We calculated the surprise for each neuron j as: <| −̃|>, where |…| denotes absolute value, and <…> denotes averaged across all 400 images presented in a single training epoch.…”