The graphical brain: Belief propagation and active inference

Abstract

This paper considers functional integration in the brain from a computational perspective. We ask what sort of neuronal message passing is mandated by active inference—and what implications this has for context-sensitive connectivity at microscopic and macroscopic levels. In particular, we formulate neuronal processing as belief propagation under deep generative models. Crucially, these models can entertain both discrete and continuous states, leading to distinct schemes for belief updating that play out on the same (neuronal) architecture. Technically, we use Forney (normal) factor graphs to elucidate the requisite message passing in terms of its form and scheduling. To accommodate mixed generative models (of discrete and continuous states), one also has to consider link nodes or factors that enable discrete and continuous representations to talk to each other. When mapping the implicit computational architecture onto neuronal connectivity, several interesting features emerge. For example, Bayesian model averaging and comparison, which link discrete and continuous states, may be implemented in thalamocortical loops. These and other considerations speak to a computational connectome that is inherently state dependent and self-organizing in ways that yield to a principled (variational) account. We conclude with simulations of reading that illustrate the implicit neuronal message passing, with a special focus on how discrete (semantic) representations inform, and are informed by, continuous (visual) sampling of the sensorium.Author SummaryThis paper considers functional integration in the brain from a computational perspective. We ask what sort of neuronal message passing is mandated by active inference—and what implications this has for context-sensitive connectivity at microscopic and macroscopic levels. In particular, we formulate neuronal processing as belief propagation under deep generative models that can entertain both discrete and continuous states. This leads to distinct schemes for belief updating that play out on the same (neuronal) architecture. Technically, we use Forney (normal) factor graphs to characterize the requisite message passing, and link this formal characterization to canonical microcircuits and extrinsic connectivity in the brain.

An active inference approach to modeling structure learning: concept learning as an example case

“…To simulate concept learning (based on the task described above) we needed 347 to specify an appropriate generative model. Once this model has been specified, one 348 can use standard (variational) message passing to simulate belief updating and 349 behavior in a biologically plausible way: for details, please see ( KJ Friston,350 FitzGerald, et al, 2017; KJ Friston, Parr, et al, 2017). In our (minimal) model, the 351 first hidden state factor included (up to) eight levels, specifying four possible types 352 of birds and four possible types of fish ( Figure 3A).…”

“…269 hidden states under two policies over three time points (i.e., two transitions), whereas in 330 the task described in this paper there are greater number of allowable policies. For more 331 information regarding the mathematics and processes illustrated in this figure, see (KJ 332 Friston, Lin, et al, 2017;KJ Friston, Parr, et al, 2017).…”

An Active Inference Approach to Dissecting Reasons for Non-Adherence to Antidepressants

“…The specific model we use is a Markov decision process (MDP; figure 1), which is particularly useful for modeling decision processes that require consideration of distal future outcomes under uncertainty (44,45). The formal basis for these models has been thoroughly detailed elsewhere (41,(46)(47)(48)(49) (also see tables 1 and 2 for more mathematical detail). Figure 1.…”

Confidence Predictions Affect Performance Confidence and Neural Preparation in Perceptual Decision Making.

“…Here, we provide important evidence that one of the multiple cues informing confidence judgments are experience-based expectations regarding the own performance. These findings can be translated into a Bayesian framework of human cognition (Ma, Beck, Latham, & Pouget, 2006;Beck et al, 2008), which proposes that continuous and flexible update of beliefs based on both existing beliefs (priors) and newly incoming information is key to efficient successful behavior (Friston, Parr, & de Vries, 2017;Hasson 2016;Ma & Jazayeri, 2014). In this model, predicted confidence acts as a prior for performance confidence and the continuous update of prior confidence allows participants to form accurate beliefs about their own performance.…”