Distinguishing imagery and perception in generative models of brain function
Year: 2026
Type: preprint
Abstract: Perception is thought to be a generative process, relying on prior beliefs and expectations to shape what we experience. This generative architecture also enables the internal generation of sensory activity in the absence of input, in the form of imagination. It remains unclear how imagery and perception can be distinguished within generative models of brain function. Here, we use Bayesian computational models to test several alternative hypotheses. We first show that imagery and perception cannot be separated within first-order Bayesian inference. Next, we show that a model which formalizes reality inferences as higher-order evaluations of sensory precision can separate imagery and perception and can explain a range of empirical observations about the relationship between perception, imagery, reality monitoring and aphantasia. We further generate novel hypotheses highlighting directions for future research. Our approach offers a simple yet powerful framework for developing a computational neuroscience of reality monitoring.
Authors: Nadine Dijkstra, Stephen M Fleming
Language: English
Subfield: Experimental and Cognitive Psychology
Field: Psychology
Domain: Social Sciences
Open Access status: green
OpenAlex ID: W7167337143