Past week, I attended and assisted at the Reinforcement Learning Summer School organized by Vincent François-Lavet. We had some great lectures on a wide variety of topics, including pure exploration in bandits, the exploration-exploitation tradeoff in RL, MCTS, symmetries and state similarities, world models and hierarchical and distributational RL.
The summer school was hosted at one the theatre halls of our university’s new building, which serves as a movie theatre at night: great seats and air quality do help with focusing for 8+ hours! A fun detail was the dramatic Indiana Jones score randomly starting at the end of the final talk by Marc G. Bellemare.
The two-year COVID-induced hiatus for in-person events did not result in less interaction. On the contrary: participants were eager to respond to questions from lecturers, pose questions themselves and share research ideas.
Personally, I was involved in assisting for the practical sessions, which proved to be a great way to get familiar with the technical content, but also to get to know the participants individually.
Additionally, I was responsible for organising a quiz on the summer school’s content near the end. This made me appreciate the talks a lot more, because I had to take notes to ensure my questions made sense. Additionally, it made me appreciate how hard it is to come up with questions that strike the right balance between clarity, challenge and fun.
The social event was (big surprise!) a boat tour of Amsterdam’s canals with lunch and drinks. We ended the week with a `pizza dinner'. These turned out to be great opportunities to talk about RL-specific topics and other interests such as music, history and experiences of PhD students in RL.