which equates to high prediction er

which equates to high prediction error. This could result in robust AONengagement for highly unfamiliar actions, as the influence of feedfor-ward/perceptual activity is heavily relied upon. When viewing an actionthat is highly familiar, however, predictions generated by the networkshould be much more precise, thus minimising prediction error. Theminimising of prediction error could also manifest as robust AONengagement, this time due to the strength of feedback signals project-ing posteriorly (which were weaker when movements were unfamiliarand prediction error was higher; see alsoCross et al., 2012). Thereciprocal nature of exchanging prediction error signals between coreAON nodes allows for the explanation of robust AON engagement forboth familiarorunfamiliar actions, relative to actions of an inter-mediate level of familiarity (illustrated inFig. 1B1). It is important tonote as well that while this Bayesian framework has been most fullydeveloped in the realm of action observation, it also has been applied toaction execution, formally known as active inference (Friston, 2005).As several authors have now suggested, a predictive coding accountof action familiarity and AON engagement could manifest as aquadratic, or U-shaped, function (Cross et al., 2012; Liew et al.,2013). However, as identified within the predictive coding literature(Kilner et al., 2007a, 2007b; Friston, 2005), a system that relies onBayesian comparisons would need to continually update predictedmovements in relation to actual movements. For example, whe