Directed base masks

Thank you so much for this amazing work!

I'm trying to understand the idea of directed base masks. I have a question: What is the difference between Gaussian Linear (GL) and Gaussian Mixture (GM), making x_i does not attend to theta_j (i != j) in GL but does in GM?  

Thanks again for your help!
