Elom Amematsro

About

Elom Amematsro is a Kempner Research Fellow at Harvard University. He earned his Ph.D. in Neurobiology and Behavior at Columbia University, where he worked with Mark Churchland and Larry Abbott at the Zuckerman Institute. His doctoral research combined large-scale neural recordings, computational modeling, and theoretical analysis to study how the motor cortex enables flexible skill learning, bridging systems neuroscience, machine learning, and cognitive theory.

Research Focus

Elom’s research at the Kempner Institute focuses on uncovering the neural and computational principles that allow humans to rapidly learn new skills by leveraging relationships to previously learned ones. He aims to identify the brain circuits that encode these relational structures and support the flexible recombination of existing skills into novel behaviors. To achieve this, his work integrates behavioral experiments, large-scale neurophysiology, and computational modeling, with an emphasis on recurrent neural networks and hierarchical learning systems. By bridging systems neuroscience with machine learning, he seeks to develop models that capture the compositional nature of human learning and apply them to artificial networks, enabling more efficient generalization and adaptability. In the long term, his goal is to use these insights not only to advance artificial intelligence, but also to inform clinical approaches for restoring cognition in individuals with neurological and psychiatric disorders.