Demonstration of proving probabilistic programs using theorem proving
Based on the theory of probabilistic relations and its implementation in Isabelle/UTP

Here we illustrate a probabilistic programming language that we are developing, its semantic implementation in Isabelle/UTP (UTP in Isabelle/HOL), the introduced Iverson bracket and distribution notations, proved algebraic laws including fixed-point theorems, reasoning about three probabilistic examples: a coin flip, the Monty Hall problem with learning facts, and a robot localisation.
The Isabelle/HOL theories for probabilistic relations can be found here.