About

Adrien Ecoffet is a highly accomplished Research Scientist with extensive experience in the field of deep reinforcement learning. He is currently employed at OpenAI, where he focuses on studying emergent behaviors in multi-agent reinforcement learning. Adrien is a co-first author of the Nature article titled "First return, then explore," which has been covered by various media outlets such as New Scientist, Scientific American, and the BBC. Before joining OpenAI, Adrien worked as a Research Scientist at Uber, where he continued his work on exploration in RL that eventually led to the Nature article. He also conducted research in AI safety, studying the implementation of results from moral philosophy in RL systems. As an AI Resident at Uber, Adrien initiated and led Go-Explore, a deep reinforcement learning algorithm that solved the grand-challenge Atari game "Montezuma's Revenge." This achievement outperformed the state of the art by a factor of 60 and led to one of the most viewed blog posts on the Uber eng blog, several news stories (MIT Tech Review, INQUIRER, VentureBeat...), and a talk at the NeurIPS 2018 Deep RL Workshop. Adrien's educational background includes a Master of Science in Computer Science from the Georgia Institute of Technology, where he focused on reinforcement learning. He also spent a semester as an exchange student at the University of California, San Diego, and holds a Bachelor of Science in Computer Engineering from Epitech - European Institute of Technology. Adrien's technical expertise includes research in reinforcement learning, AI, and multi-agent systems. He has 8.45 years of relevant experience in the field and is highly regarded for his contributions to the development of deep reinforcement learning algorithms. Adrien is a passionate researcher who is constantly seeking new ways to advance the field of AI and its applications.