The following has been transcribed and edited for clarity by Sean O'Sullivan. SharonMarcus (SM): I'm Sharon Marcus-I'm a Professor of English at Columbia University: I'm a Victorianist, and I work on nineteenth-century French literature, so I'm well-acquainted with the history of seriality. I'm also the Dean of Humanities and Editor of publicbooks.org. I would like to thank Lauren Goodlad and Sean O'Sullivan and EileenGillooly and everyone at the Heyman Center for putting this on today. I am going to introduce our panelists, although they don't really need an introduction. They say of great actors that you would be happy to listen to them read the phone book. I think we can say of our panelists that we' d be happy to hear them write a review of the phone book, write a novel based on the phone book, or produce the phone book as a radio podcast.[laughter] But rituals are important, so here we go: Lev Grossman is the author of five novels, including the #1 New York Times-bestselling Magicians trilogy, which is now an

No abstract

Deep reinforcement learning (DRL) is one of the most powerful tools for synthesizing complex robotic behaviors. But training DRL models is incredibly compute and memory intensive, requiring large training datasets and replay buffers to achieve performant results. This poses a challenge for the next generation of field robots that will need to learn on the edge to adapt to their environment. In this paper, we begin to address this issue through observation space quantization. We evaluate our approach using four simulated robot locomotion tasks and two state-of-the-art DRL algorithms, the on-policy Proximal Policy Optimization (PPO) and off-policy Soft Actor-Critic (SAC) and find that observation space quantization reduces overall memory costs by as much as 4.2× without impacting learning performance.

No abstract