Fig. 5: Estimated learning curves and firing rates of example changing cells.
From: The contributions of entorhinal cortex and hippocampus to error driven learning
The estimated trial-by-trial firing rates (purple lines) superimposed on estimated learning curves (teal lines) of an example sustained changing cell in HPC was plotted in a, a baseline sustained changing cell in b, an example sustained changing cell in EC in c, and a baseline sustained changing cell in d. The blue-filled circles on top of each subplot indicate correct trials and the red asterisks indicate the error trials. The 95% confidence bounds of estimated learning curves were plotted in teal dashed lines. The trial-by-trial firing rates plotted here were estimated with the same state-space algorithm used to estimate behavioral performance33 to avoid the difficulty of recognizing the correlation between the firing rates and the behavior performances due to the noisy neural spiking with inherited variability. Performances are shown in Supplementary Data 1.
