Juji is an artificially intelligent chatbot-based assistant that aims to help teachers engage with students, at scale and in a personalized way. The idea is to free up more time for teachers, and admin staff, to focus on other tasks. This is a complete platform, so it is a chatbot AI builder as well as the front-end system itself. So schools and, primarily universities and colleges, can work on their personalized AI to be used at their educational institution. This can range from helping with student recruitment to guiding students in a course.

Some few weeks ago I posted a tweet on "the most common neural net mistakes", listing a few common gotchas related to training neural nets. The tweet got quite a bit more engagement than I anticipated (including a webinar:)). Clearly, a lot of people have personally encountered the large gap between "here is how a convolutional layer works" and "our convnet achieves state of the art results". So I thought it could be fun to brush off my dusty blog to expand my tweet to the long form that this topic deserves. However, instead of going into an enumeration of more common errors or fleshing them out, I wanted to dig a bit deeper and talk about how one can avoid making these errors altogether (or fix them very fast).

In this post, I'm going to cover tricks and best practices for how to write the most effective reward functions for reinforcement learning models. If you're unfamiliar with deep reinforcement learning, you can learn more about it here before jumping into the post below. Crafting reward functions for reinforcement learning models is not easy. It's not easy for the same reason that crafting incentive plans for employees is not easy. We get things affectionately known as the cobra effect.