# Efficient Gradient-Based Inference through Transformations between Bayes Nets and Neural Nets

arXiv: https://arxiv.org/abs/1402.0480 · HTML: https://mlsanity.com/paper/1402.0480/

## Key facts

- Cited by 5 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 3 Feb 2014.

## Built on by (most cited first, 5 of 5)

- [A Deep Latent Space Model for Graph Representation Learning](https://mlsanity.com/paper/2106.11721/) (2021)
- [Meta-learning Amidst Heterogeneity and Ambiguity](https://mlsanity.com/paper/2107.02228/) (2021)
- [Embedded Topic Models Enhanced by Wikification](https://mlsanity.com/paper/2410.02441/) (2024)
- [Stick-Breaking Mixture Normalizing Flows with Component-Wise Tail Adaptation for Variational Inference](https://mlsanity.com/paper/2510.07965/) (2025)
- [SGD for Variational Inference: Tackling Unbounded Variance via Preconditioning and Dynamic Batching](https://mlsanity.com/paper/2605.07531/) (2026)

## Abstract (arXiv)

Hierarchical Bayesian networks and neural networks with stochastic hidden units are commonly perceived as two separate types of models. We show that either of these types of models can often be transformed into an instance of the other, by switching between centered and differentiable non-centered parameterizations of the latent variables. The choice of parameterization greatly influences the efficiency of gradient-based posterior inference; we show that they are often complementary to eachother, we clarify when each parameterization is preferred and show how inference can be made robust. In the non-centered form, a simple Monte Carlo estimator of the marginal likelihood can be used for learning the parameters. Theoretical results are supported by experiments.