SingleR diagnostics about the annotation process

[BioRB]

Dear Support,
in the original SingleR we can take advantage of a number of helpful diagnostics about the annotation process that can be used to determine whether the assignments are appropriate.
The most obvious diagnostic reported by SingleR() is the nested matrix of per-cell scores in the scores field.

pred.grun$scores[1:10,]

where or how I can find this scores in Stereopy - SingleR?