Towards alleviating catastrophic forgetting for deep RL multi-task learning.
Check out the work in this repo!