Charlie Steiner comments on Conditioning Predictive Models: Interactions with other approaches