Skip to main content
. 2010 Jun 24;66(6):937–948. doi: 10.1016/j.neuron.2010.05.018

Figure 3.

Figure 3

Spike Rate across Different Stimulus Distributions

(A) Example baseline rate-ILD function of one neuron in the IC and histograms of three ILD distributions with different means. Each histogram shows the distribution of the 90,000 ILDs presented during all unique adaptation period sequences for a particular mean.

(B) Average firing rate as a function of time (smoothed) from the same neuron for ILD distributions with different means, averaged over 90 unique adaptation periods. Dashed lines indicate expected adaptation rates based on the overlap between the ILD distribution and the baseline rate-ILD function (see Experimental Procedures for details on calculation of expected rates).

(C) Observed adaptation rates during the baseline (zero mean) ILD distribution plotted against the observed (circles) and expected (crosses) adaptation rates for the distributions with mean ILDs of either −15 dB (green) or +15 dB (red). A circle lying on the black dotted unity line indicates that the adaptation rate remained constant across a change in mean.

(D) Mean observed (black) and expected (gray) adaptation rate for distributions with different means.

(E) Example baseline rate-ILD function for the same neuron whose response is shown in (A) and (B), together with histograms of stimulus distributions with two different variances.

(F) Average firing rate as a function of time (smoothed) from this neuron for distributions with different variances. Dashed line indicates expected adaptation rate based on the overlap between the low-variance distribution and the baseline rate-ILD function.

(G) Observed adaptation rates for the high- (SD = 20 dB) variance distribution versus observed (circles) and expected (crosses) adaptation rates for a low- (SD = 10 dB) variance distribution. A circle lying on the dotted black unity line indicates that the adaptation rate remained constant across a change in variance.

(H) Mean observed (black) and expected (gray) adaptation rates for distributions with different variances. Error bars are ± SEM.