Bayesian transfer in a complex spatial localisation task

Abstract

7Prior knowledge can help observers in various situations. Adults can simultaneously 8 learn two location priors and integrate these with sensory information to locate 9 hidden objects. Importantly, observers weight prior and sensory (likelihood) 10 information differently depending on their respective reliabilities, in line with 11 principles of Bayesian inference. Yet, there is limited evidence that observers 12 actually perform Bayesian inference, rather than a heuristic, such as forming a look-13 up table. To distinguish these possibilities, we ask whether previously-learnt priors 14 will be immediately integrated with a new, untrained likelihood. If observers use 15

Where did I park my car? Influence of landmark permanency on navigation

“…The next step would be to test the capacity of transferring the learned to other situations. An additional evidence that this probabilistic learning follows Bayesian principles would be to show that the prior of permanency, learned in a context of homing task navigation, is transferable to another type of task, since the capacity of transfer is a signature of the Bayesian principles (Kiryakova et al, 2019;Maloney & Mamassian, 2009). To conclude, navigation is one of the most fundamental abilities in animals including humans.…”

“…Our study is embedded in the Bayesian Decision Theory applied to navigational behavior. Some studies already investigated navigational behavior with the lens of the Bayesian theory (Cheng et al, 2007;Kiryakova et al, 2019;Nardini et al, 2008;Pfuhl et al, 2011). This theory provides a framework for a better understanding of navigational behavior.…”

“…The theory predicts that a decision is based on the combination of priors and sensory information, where the weight allocated to one or the other depends on their reliabilities, assessed by the variance of their distributions; the higher the variance, the lower the allocated weight (Ernst & Bülthoff, 2004;Maloney & Mamassian, 2009). While the theory has been mainly applied to model perception, here we apply the mechanisms to navigational behavior (see also Cheng, Shettleworth, Huttenlocher, & Rieser, 2007;Kiryakova, Aston, Beierholm, & Nardini, 2019;Nardini, Jones, Bedford, & Braddick, 2008). The probability that we have experienced a church moving is much lower than that of experiencing a moving truck.…”