ABSTRACT
In recent decades, Bayesian modeling has achieved extraordinary success within perceptual psychology. A natural question raised by Bayesian perceptual psychology is how the brain implements Bayesian inference. How do neural states physically realize the prior probability, the prior likelihood, and the posterior? Which neural operations effectuate the transition from priors to posterior? These questions have been intensively studied in computational neuroscience, and there are now several proposed neural implementation mechanisms. I canvass some proposed implementation mechanisms, including both predictive coding and alternatives. I then draw morals regarding the debate between realist and instrumentalist interpretations of Bayesian perceptual psychology.
