Jonathan Pillow, Associate Professor of Psychology and the Princeton Neuroscience Institute
Pillow earned a dual bachelor’s degree in mathematics and philosophy from the University of Arizona in 1997. He then earned his doctoral degree from New York University’s (NYU) Center for Neural Science in 2005. His thesis was titled, “Neural coding and the statistical modeling of neuronal responses.”
He was also a postdoc at the Gatsby Computational Neuroscience Unit, University College London. Before coming to Princeton, Pillow was an assistant professor at The University of Texas at Austin. Pillow has been at Princeton since 2014.
In addition to his more than 80 published research papers, Pillow has won grants from the National Institute of Health, Simon Foundation, National Science Foundation, and other organizations. He has earned numerous awards over the years, including the Sloan Research Fellowship and the Presidential Early Career Award for Scientists and Engineers (PECASE).
Pillow’s main focus of research is concerned with analyzing brain activity and its connection with human behavior by using statistics, machine learning, and computational methods.
“What’s exciting about neuroscience is there are a variety of powerful new technologies for recording neural activity in the brain. When recording neurons in the brain a generation ago, you could only record one neuron at a time. Now you can record hundreds to thousands of neurons simultaneously,” Pillow said. “There this is a great need now for advanced statistical methods to make sense of all this data, because meaningful signals are distributed across large populations noisy neurons.”
Pillow’s research specifically examines how neurons respond to sensory stimuli and seeks to determine what aspects of neural activity carry information. Neurons can respond to the same stimulus in different ways depending on the state of the test subject. Pillow’s lab develops statistical models to understand this state-dependent neural activity.
“For example, we want to understand how the brain makes decisions,” he said. “When you make a decision, some neurons may gradually increase their firing rate, reflecting an increase in evidence or confidence. Other neurons, on the other hand, may jump suddenly from one firing rate to another, reflecting the decision itself.”
These questions are important, Pillow said, because they answer basic questions about brain function, such as the mechanisms underlying perception, cognition, memory, and movement. By understanding these functions, researchers will gain insight in how to understand and cure brain disorders such as depression, dementia, Parkinson’s, etc.
On machine learning and artificial intelligence technologies, Pillow said the discipline and CSML’s mission are vital because the tools being taught via the center provide deep insight into many questions across the sciences.
“Machine learning is about to revolutionize how we understand information processing in both natural and artificial systems, and CSML is at the heart of these efforts,” he said. “CSML provides a curriculum that students need to tackle a wide variety of important problems in science and society. Science in particular is undergoing a fundamental revolution that requires researchers with the ability to manipulate and analyze large datasets that modern technologies are allowing us to collect.”