The Invaliant Causal Prediction Paper

At the time, i knew that leon was very excited about galsality. We all had this vague intuition that if you know that x causes why, then you change the distribution of exs and you know what's going to happen. Idea basically that for out of distribution generalization, if you can understand what causes the differences between th descriptions, thenyo can generalize better. So after the gunpaper came by, i went to montreal for two weeks with ishan who came up with his improse basstin ganari.