[Wien] LDA+U results for CeO2 don't look right

[Craig P Plaisance]

I did a calculation on CeO2 using the PBE GGA with LDA+U(AMF) without SO
coupling.  I am using U=0.74 and J=0.07.  When I looked at the results, the
occupancy of the Ce(4f) states was 0.4 for each spin.  Shouldn't the strong
coulomb repulsion between f electrons prevent the other spin state from
being occupied?  Am I doing something wrong, or is this the best that LDA+U
can do?

Thanks,

Craig Plaisance