(A) Schematic showing the different stages of processing in the electrosensory system. Recordings were made from single (N = 18) as well as pairs (N = 8) of afferents. (B) Example spike trains (blue, top), firing rates averaged across trials (middle, blue), and correlation coefficient (purple) from an example afferent pair to different stimulus waveforms associated with the same electrosensory object. Note that each stimulus was presented at least 20 times. The horizontal bars (shaded gray) represent the chirp window (30 ms) used for evaluation. (C) Population-averaged firing rate responses to the different stimulus waveforms during the 30 ms chirp window shown in B. (D) Population-averaged tuning curves obtained from single neuron firing rates (blue) and from the correlation coefficient (purple). '+ chirps' and '- chirps' were defined as the stimulus waveforms that gave rise to positive (i.e. increases in firing rate) and negative (i.e. decreases in firing rate) responses in afferents, respectively. (E) Correlation coefficient as a function of time in response to all stimulus waveforms. Inset: Invariance score computed from single afferents (left), from the population-averaged activity (center), and from the correlation coefficient (right). (F) Invariance score computed from single afferent activity (blue) and from correlated activity (purple) computed for all stimulus waveforms, '++/- - chirps', and '+- chirps'. '*' indicates statistical significance at the p = 0.05 level using a one-way ANOVA with Bonferroni correction.