I am attempting to calculate predictions after a melogit model with and without best linear unbiased predictors. In SAS I would do this as follows:

Code:
PROC GLIMMIX DATA=&ANALYSIS NOCLPRINT MAXLMMUPDATE=100;
CLASS PROVID;
ODS OUTPUT PARAMETERESTIMATES=&EST (KEEP=EFFECT ESTIMATE STDERR);
MODEL OUTCOME=&MODEL_VAR
        /D=B LINK=LOGIT SOLUTION;
XBETA=_XBETA_;
LINP=_LINP_;
RANDOM INTERCEPT/SUBJECT=PROVID SOLUTION;
OUTPUT OUT=OUTCOME
        PRED(BLUP ILINK)=PREDPROB PRED(NOBLUP ILINK)=EXPPROB;
For SAS the options above are:
BLUP=linear predictor
NOBLUP = marginal linear predictor
ILINK = predicted mean

However, in Stata I am not totally sure how to do this. Here is the code I am using. My understanding is that the predict command after running an melogit predicts the BLUP at the means. But I am not sure if the xb option would be calculating predictions at the means without the BLUP.

Code:
melogit outcome modelvars  || secondlevel:
predict predicted
predict expected, xb