Madelyn Andersen | arXiv Science

arXiv2026

The Curious Case of the Default Settings: Evaluating Default Performance of Variational Inference Software

We systematically evaluate a suite of off-the-shelf variational inference (VI) software packages from the perspective of a standard practitioner. Using simple analytic benchmark models, we assess the accuracy and stability of the default VI settings in PyMC, NumPyro, and TensorFlow Probability. Unlike previous research focusing on methodological advances, our evaluation emphasizes software implementation and the default configurations that typical users encounter. Our results show that default settings can yield biased approximations of posterior summaries even for simple one-dimensional conjugate models, controls of initialization and transformations differ between software implementations, and relying on defaults may yield silent failures or poor approximations.

stat.CO↗

arXiv2021

Finite rank kernel varieties: A variant of Hilbert's Nullstellensatz for graphons and applications to Hadamard matrices

Graphons are symmetric measurable functions that arise from a sequence of graphs. A graphon variety is the a set of all graphons defined by a condition of the form $t(g, W) = 0$ for a fixed quantum graph $g$, where $t(.,.)$ is the homomorphism density and a quantum graph is a formal linear combination of multigraphs. Using a method of representing graphs as polynomials, we construct an epimorphism from the space of quantum graphs to a subring of the complex polynomial ring that is invariant under permutations of variables. When graphons are of finite rank, we demonstrate that an analog of the "ideal" inverse in Algebraic Geometry is an ideal in our polynomial representation. Defining an algebraic kernel set using kernel varieties, we demonstrate that we can call such sets closed under the Zariski Topology. We determine several ties to Algebraic Geometry as a result of utilizing finite rank kernels and discover that a weaker version of Hilbert's Nullstellensatz applies to kernel zero-sets with respect to homomorphism density. Throughout, we examine the connection between Algebraic Geometry and Graphon Theory.

math.AG↗