Dylan Losey, associate professor in mechanical engineering, won best paper at the Conference on Robot Learning held virtually November 16-18, 2020. The conference is the flagship conference for work at the intersection of robotics and machine learning. Losey shared authorship with faculty at Stanford University.

The paper is titled "Learning Latent Representations to Influence Multi-Agent Interaction." According to Losey: "It's hard for robots to seamlessly interact with humans, because humans often change their behavior in response to the robot. In this paper, we explore how robots can co-adapt alongside humans and other intelligent agents. Our proposal is a learning approach where the robot captures the human's behavior as a high-level strategy, and then influences that strategy over time."