Alex Anderson

Research Scientist


Olshausen Lab

Current Research

My research is energized by several key themes: machine learning inspired by neuroscience, computation with noisy elements, and using probabilistic graphical models to structure neural network computation. My thesis work develops a neural model that demonstrates the benefits of eye motions for high-acuity vision. My machine learning projects involve: binary neural networks, recurrent neural networks for noisy sequence prediction, artistic stylization of movies, and hierarchical neural network models that drive cars.


My passion for learning began when I discovered the beauty of Euclidean geometry as a young freshman in high school, and was cultivated by the study of the elegant mathematics of high school math competitions. I loved creating clever math problems. Unexpectedly, this played a large role in my admission to Washington University in St. Louis where I studied mathematics and physics. The one idea that inspired me most in college was how one could use mathematics and careful experimentation to understand the intuition-defying mysteries of quantum mechanics and special relativity. A close second was the elegance and complexity of the biological process of going from DNA to RNA to proteins. Consequently, I joined the physics department at UC Berkeley as a doctoral student. After brief stints in quantum computing and biophysics, I joined Bruno Olshausen’s lab in the Redwood Center for Theoretical Neuroscience where I study computational neuroscience and machine learning. In my free time, I love all types of exercise (from weight lifting and half-marathon running to dancing and yoga), I enjoy learning about psychology and entrepreneurship, and I am energized by progressive politics.