Skip to main content
. 2016 Dec 27;114(2):388–393. doi: 10.1073/pnas.1614763114

Fig. S1.

Fig. S1.

Correlations between neural data and hidden layers 1 (H1) and 4 (H4) of a deep neural network trained on our stimuli, as a function of time (milliseconds). Separate plots are given for each region of interest (rLO-faces, rFG-faces, and lV1). The horizontal black line indicates postbaseline time points at which correlations differed significantly (FDR < 0.05) between H1 and H4.