Scheduling pipeline - do the variable values become hard coded?

Summary:

Does a scheduled pipeline only ever remember what the variables were at the point in time it was scheduled, or does it refer to the pipeline settings each run and pick up any changed variables?

Content (please ensure you mask any confidential information):

I have a couple of Pipelines, in which I have set the variable values at Pipeline level for STARTPERIOD and ENDPERIOD. Naturally those get pulled through to the relevant stages of the Pipeline in the usual way using $STARTPERIOD and $ENDPERIOD.

When I started my schedules for the pipeline the ENDPERIOD value was (and still is) set to Nov-23. In due course I will be updating the pipeline to make it Dec-23.