To me, this doesn’t really look like the little sample is trying to levitate. It looks like it’s trying to rotate, almost like it’s trying to line up with the m

To me, this doesn’t really look like the little sample is trying to levitate. It looks like it’s trying to rotate, almost like it’s trying to line up with the magnetic field. Magnetic dipoles in a magnetic field experience a torque, and if a little sliver of stuff experiences a torque trying to make it point a particular direction, this could easily overcome the force pulling it toward the magnet. (The torque on a dipole is proportional to the applied field. The force is proportional to the gradient of the field. And the experimenters seem to be using a pretty big flat magnet or a big solenoid, close to the center, which is likely to have a fairly spatially constant field near the surface.) I’m reminded of once holding a bar magnet in a small plasma testbed electromagnet. Small = a few feet long and maybe 2 feet across. (The big one across the room was about the size of a semi truck IIRC.). I naively expected the magnet to be pulled in toward the center. Nope, not so much, but it tried quite hard to twist itself in my hands.