Leer is a tool for learning more about various Reinforcement Learning algorithms such as value iteration or prioritised sweeping. Leer allows the you to create a maze and then pits the algorithms against it. While solving the maze, Leer provides a graphical representation of the algorithm's process through the maze.
The focus of this application is to demystify Reinforcement Learning algorithms, making them accessible to people who want to get a handle on the internal workings of the algorithms. A choice of four algorithms (value iteration, q-learning, poly iteration and prioritised sweeping) means that you can compare the implementations and how they achieve a solution.