>>528105940
how about vanilla-age regression ge?