We first looked at behavior - comparing similar to dissimilar speed profiles - and found that there was no difference in drift between these sets of data. Place cells drifted at the same rate in both sets of behavior, ruling out behavior as the driver of drift.