Strategic stability under regularized learning in games