Sometimes refers to the behaviour of cohomology under base change between separably closed fields. I think this is true for etale and unramified cohomology. Suslin proved something for K-theory with coefficients, and Lecomte something for Chow groups.
nLab page on Rigidity