Hello, I'm a bit confused about the note referring to SparseVariationalGP model, since I can't find such a model in Pyro. On the other hand, if this is just a typo, and the intention was to refer to VariationalSparseGP I have two remarks

1. Based on references (Scalable variational Gaussian process classification, James Hensman, Alexander G. de G. Matthews, Zoubin Ghahramani), as I understand it, the model represents the Sparse Variational Gaussian Process (SVGP). If this is the case, the name of the class VariationalSparseGP seems a bit counterintuitive (SVGP vs VSGP) since the majority of the literature recognizes this model under the SVGP abbreviation.
2. What is meant by VFE being equivalent to SVGP? I thought that SVGP is an extended/improved version of VFE, see [1, 2].

Thank you for any help!