Building on previous work (Metzen et al., 2016), a combination of neurophysiological and behavioral approaches reveals that changes in the background strongly impacts invariant coding and perception of behaviourally relevant signals.

Whether central vestibular neurons implement faithful stimulus encoding for the vestibulo-occular reflex or optimized coding via temporal whitening for other vestibular functions is determined by neural variability.

A compact neural network can form various associative memories that are encoded in a distributed manner, where each neuron stores different components of the memory.

Neurophysiological and behavioral approaches reveal how coordinated input from descending pathways shapes the tuning properties of electrosensory neurons in order to optimize coding of natural stimuli through temporal whitening.

Neurons in the brain exhibit activity with various relations to locomotion and these signals are best decoded by combining activities from many neurons.

Neural oscillations are a necessary consequence of efficient coding of sensory signals by a spiking neural network, limited by synaptic delays and noise.

Neural circuits in weakly electric fish perform a set of computations to allow natural communication signals to be perceived independently of their context.

Separating behaviorally relevant signals from irrelevant signals reveals that neural signals previously considered to contain little information encode rich information and suggests that linear readout may be performed in motor cortex.

Computer simulations of interaural time difference decoders show that heterogeneous tuning of binaural neurons leads to accurate sound localization in natural environments.

The prefrontal cortex encodes both stable and dynamic representations of expected value, providing mechanisms to support robust as well as flexible access to value information during temporal delays.