Edward provides a testbed for rapid experimentation and research with probabilistic models. Here we demonstrate examples from each of the three components defining Edward: **Models**, **Inference**, and **Criticism**. In the **End-to-End** section, we demonstrate examples that sweep through the whole process.

- Probability models
- Bayesian linear regression
- Mixture of Gaussians
- Gaussian process classification
- Bayesian neural network
- Probabilistic decoder

- Inference of probability models
- Variational inference
- \(\text{KL}(q\|p)\) minimization
- \(\text{KL}(p\|q)\) minimization
- Maximum a posteriori estimation
- Inference networks