Inverse prediction is the opposite of prediction. It is the prediction of x values from given y values. But in logistic regression, instead of a y value, you have the probability attributed to one of the Y levels. This feature works only for two-level nominal responses.
The Fit Model platform also has an option that gives an inverse prediction with confidence limits. Inverse Prediction in the Fitting Linear Models book gives more information about inverse prediction.
Related Information 

Help created on 9/19/2017