Abstract: In this workshop, we will see how it’s possible to use Reinforcement Learning to solve real-world problems. Reinforcement Learning recently had incredible progress in the industry and established itself as one of the best techniques for sequential decision-making and control policies. In particular, we will see how Reinforcement Learning with Human feedback and other techniques improved the performance of LLMs.
We are now just scratching the surface of its applications.
In this session, we will start by exploring its fundamentals, like Q-learning, and we will end with Deep Reinforcement Learning algorithms, like Actor-Critic models and how to use these techniques to improve LLMs.
Learning outcomes
The theory behind the RL algorithms and the difference between the various methods.
The advantages and disadvantages of each model and how they can be successfully applied in different scenarios.
How to apply RL techniques in different scenarios, including improving LLMs.
How to code some of these algorithms in Python
Bio: Leonardo De Marchi holds a Master degree in Artificial intelligence and has worked as a Data Scientist in the sports world, with clients such as the New York Knicks and Manchester United. He now works in Thomson Reuters as VP of Labs and also provides consultancy and training for small and large companies. His previous experience includes being Head of Data Science and Analytics in Bumble, the largest dating site with over 500 million users, heading the team through acquisition and an IPO.



