Predictive coding: the idea that the brain generates hypotheses about the possible causes of forthcoming sensory events and that these hypotheses are compared with incoming sensory information. The difference between topdown expectation and incoming sensory inputs, that is, prediction error, is propagated forward throughout the cortical hierarchy. Predictive timing (temporal expectations): an extension of the notion of predictive coding to the exploitation of temporal regularities (such as a beat) or associative contingencies (for instance, temporal relation between two inputs) to infer the occurrence of future sensory events. Top-down processing: efferent neural operations that convey the internal goals or states of the observer. This notion generally includes different cognitive processes, such as attention and expectations (Box 1).
Neural oscillations: neurophysiological electromagnetic signals [from LocalField Potentials (LFP), electroencephalographic (EEG) and magnetoencephalographic recordings (MEG)] that reflect coherent neuronal population behavior at different spatial scales. These signals have been labeled as a function of their frequency from human surface EEG: delta (2-4 Hz), theta (4-8 Hz), alpha (8)(9)(10)(11)(12), and gamma bands (30-100 Hz). The mechanistic properties of oscillations are computationally interesting as a means of explaining various aspects of perception and cognition, for example, longdistance communication across brain regions, unification of various attributes of the same object, segmentation of the sensory input, memory etc.