Can working memory be explained by predictive coding?

Predictive coding (PC) is a theory in cognitive/computational neuroscience which explains cortical functions with a hierarchical process of minimizing prediction errors. PC provides a neuronal scheme (through neurons or neural populations) for implementing Bayesian inference in the brain. PC recovers the hidden state of the world from sensory input (passive inference) and selects actions to reach the goals the agent has (active inference). Since its discovery, PC has been found to be a unifying theory explaining more and more cognitive functions, including perception, attention, and action planning. In this paper, we review and discuss how PC can be used also as a powerful tool to understand working memory (WM), an essential function for executive control. Specifically, we sought answers in the literature to the following questions: 1. How is WM maintained and updated? 2. What is the relationship between attention and WM and how do they interact? 3. Why does WM have limited capacity? and 4. Why is WM hierarchical? Modelling WM in PC frameworks provides alternative explanations to some long-standing questions about WM and may help with resolving the conflicts between WM theories. We expect such alternative explanations can help future researchers, such as in developing computational tools to improve treatments for brain disorders and more robust artificial working memory in artificial intelligence.