Habitual and Reflective Control in Hierarchical Predictive Coding

TL;DR

This paper proposes that Hierarchical Predictive Coding (HPC) can unify habitual and reflective behaviors within a single framework, explaining how different levels of processing are engaged depending on the task's complexity.

Contribution

It demonstrates that HPC can account for both habitual and deliberative behaviors as a continuum, with hierarchical layers engaged dynamically based on task demands.

Findings

HPC can explain both habitual and reflective control as a continuum.

Higher layers of HPC are recruited only for complex, deliberative actions.

Learning is distributed across the hierarchy, with selective layer engagement.

Abstract

In cognitive science, behaviour is often separated into two types. Reflexive control is habitual and immediate, whereas reflective is deliberative and time consuming. We examine the argument that Hierarchical Predictive Coding (HPC) can explain both types of behaviour as a continuum operating across a multi-layered network, removing the need for separate circuits in the brain. On this view, "fast" actions may be triggered using only the lower layers of the HPC schema, whereas more deliberative actions need higher layers. We demonstrate that HPC can distribute learning throughout its hierarchy, with higher layers called into use only as required.