Could indicate some superconducting regions in the whole sample, or some weird anisotropic superconduction at higher temps...

That sharp drop, if not equipment error, could be a lot of things.