Fetching the paper…

Multi-Pass Q-Networks for Deep Reinforcement Learning with Parameterised Action Spaces · Around