I have a similar issue with VESTA 3.5.8 on Linux, though for a .cif file. This happens after I open a unit cell of graphene, transform it to replicate it and add atoms at all site positions, then add an adatom on top of the graphene, and then save it as a cif file.
The weird part is that the cif file contains only the right number of atoms, and if I open the file with another software, I only see the supercell I created. In VESTA, on the other hand, it extends on the x and y planes a number of times.