Okay, I was confused by the CSV file which contained the value of the TSF, so I assumed that the plot was also the TSF.
New question: did you only decrease the TSF to 0 and did nothing else? Or did you also command the system to Off?
The step in TargetPosition could then be explained: when you switch off the drive, we sychronize the setpoint to the actual position automatically, to prevent the axis from jumping if you move it when it is turned off and then turn it back on. The target is always a bit different from the actual, so that is why you see the little jump in the setpoint.