The general answer for pull up and pull down on input pins are: Input pins always need to be in a defined (electric) state, otherwise they are going to float (and fluctuate between sensing High and Low).
Your hardware design (buttons etc.) can be using either High or Low as the 'Active state', like in 'When the button is pushed' then the button push connects the input pin to either High or Low. And your pull (up or down) needs to be the oppesit of the Active state.
Or in short: you use pull down (on an input pin) when the button (or other active state signal) is High.
It is a common practice to use the non-active state (the 'resting state' or 'no active input right now state') to be High, and then to let the active state be Low - because it is an easy way to eliminate most types of 'false' input signals. Most 'false' signals happen because of static electricity or because of other 'electro magnetic influences', and these influences will typically induce electricity into the wire. But if the resting state iof the wire/input is already 'I am full of electricity' then a little more voltage makes no change to the High state of the resting wire.
And thus, most MCU's have an option to activate a pull-up resistor, and thus use the High state as the inactive state, and the Low state as the Active state for signals.
But sometimes (because off.... ?? .... someone designed a sensor that way?!) it is necessary to use the Low state as the inactive state, and in those situations one would need a pull-down resistor to define the inactive state of the wire/input.
Does this answer your question?
