Dissociation Between The Use of Implicit and Explicit Priors in Bayesian Perceptual Inference
Caroline Bévalot, Atomic Energy Commission, National Institute of Health and Medical Research, University Paris-Saclay & Sorbonne, France; Florent Meyniel, Atomic Energy Commission, National Institute of Health and Medical Research, University Paris-Saclay, France
Posters 1 Poster
Pacific Ballroom H-O
Thu, 25 Aug, 19:30 - 21:30 Pacific Time (UTC -8)
Our brain constantly uses prior knowledge that reflects the statistics of our environment to shape our perception. Those statistics can be implicit, not directly observable but learned from observations, or explicit, communicated directly to the observer, especially in humans. Those different origins and mechanisms for acquiring priors may influence perception differently. Here, we manipulated the strength of priors and sensory likelihood to study perceptual inference in both implicit and explicit contexts. Using Bayesian models of learning and decision, we showed that subjects performed worse in the explicit than implicit context because they neglected more the sensory likelihood. The weight of the likelihood was highly correlated between contexts (but different on average) across individuals, but the weight of priors was unrelated. Those results support a dissociation in perceptual inference between the use of implicit vs. explicit priors. Many previous studies reported suboptimality of perceptual inference in healthy subjects or psychiatric disorders; those results could be reinterpreted in light of the implicit-explicit origin of priors.