Topic 1: Looking beyond MDPs, why are they limiting (Doina)
Topic 3: RL papers need a shared theoretical framework (Aditya)
Topic 4: How to make experience ultra reusable? (Glen)
Topic 6: Multi-agent reinforcement learning and continual learning (Sarath)
Topic 7: Model-based / offline RL: amortized Bayesian agents with causal compositional knowledge & inference (Yoshua)
Topic 8: Data and computational efficiency (Aaron)