Ida Momennejad

About

I am a Senior Researcher in Reinforcement Learning at Microsoft Research, New York City. Previously I was as an Associate Researcher in the Department of Biomedical Engineering at Columbia University, and before that a postdoc in computational neuroscience at Princeton University. I completed my PhD at Humboldt University in Berlin, Germany where I did my research at the Bernstein Center for Computational Neuroscience. I received my MSc degree in Philosophy of Science in Utrecht, the Netherlands, and my undergraduate degree in Software Engineering in Tehran, Iran.

My research interests are at the intersection of reinforcement learning and human cognition & behavior. I am excited about RL approaches to memory and planning. I’ve studied how humans build models of the environment and how this learning interacts with memory, exploration, & planning in both individual and multi-agent settings. I’ve tested some of these models using behavioral experiments, fMRI, and electrophysiology. Some recent projects I’m excited about include building more efficient models of memory, multi-agent simulations and metalearning, computational approaches to mental health, and RL agents with human-like navigation in games.

For a list of preprints and papers, please visit my personal webpage (opens in new tab).

If you prefer podcasts, BrainInspired (opens in new tab) & Parsing Science (opens in new tab) have kindly featured my work.
If you like scientific debates about biological & artificial learning by neuroscientists, philosophers, and AI researchers, every Friday (4 PM ET) I cohost The Learning Salon (opens in new tab) (with Joshua Vogelstein and John Krakauer).