Valiant uses computer science to learn intricacies of brain

Leslie Valiant, a renowned neuroscientist and computer scientist at Harvard University, presented the Cox Distinguished Lecture

Scientists have been studying the brain for hundreds of years, and while there have been many discoveries, progress is slow because of the brain's incredible complexity.
Aaron Bobick, Leslie Valiant and Jerry Cox
Aaron Bobick, Leslie Valiant & Jerome Cox

Leslie Valiant, a renowned neuroscientist and computer scientist at Harvard University, acknowledges that progress is slow, but sees his research as a way for computer science to contribute to neuroscience. As the T. Jefferson Coolidge Professor of Computer Science & Applied Mathematics at Harvard, Valiant presented the Cox Distinguished Lecture at the School of Engineering & Applied Science Oct. 6.

Valiant's research looks at how many neurons the brain uses to perform tasks, to store things in memory, to retrieve memories and how those neurons communicate with each other. For example, when we meet a new person, how many neurons does the brain use to learn the person's name and remember it — is it one, 10 or millions?

His interest is in symbolic processing, in which the brain is presented with information that it is able to process, then make associations with those words. The brain has to perform this hundreds of thousands of times over a person's lifetime without compromising earlier memories. He also explained chunking, in which the brain uses short-term memory more efficiently by grouping information, such as remembering first and last names together.

Valiant builds computational models to explain how our brain does operations, such as making associations or inductive learning.

"You know what green means, you know what fox means, and if I tell you there's a restaurant called the Green Fox, it's not too hard to remember," he said. "What's impressive isn't that we can do this on a small scale. What's impressive is that we can do this over a lifetime and hundreds of thousands on top of each other, which does not compromise your early memories."

The McKelvey School of Engineering at Washington University in St. Louis promotes independent inquiry and education with an emphasis on scientific excellence, innovation and collaboration without boundaries. McKelvey Engineering has top-ranked research and graduate programs across departments, particularly in biomedical engineering, environmental engineering and computing, and has one of the most selective undergraduate programs in the country. With 140 full-time faculty, 1,387 undergraduate students, 1,448 graduate students and 21,000 living alumni, we are working to solve some of society’s greatest challenges; to prepare students to become leaders and innovate throughout their careers; and to be a catalyst of economic development for the St. Louis region and beyond.