top of page

How is cognition implemented?

Biological brains endow organisms with the power of cognition. We don't understand how they achieve this seeming miracle.

This power wasn't acquired all at once. Brains evolved over hundreds of millions of years, all the while providing animals with a powerful real-time control system to gather information from the world, interpret it, and act on it.

We hypothesize that the real-time production of behavior, by the complex dynamical system that is every brain, forms the foundation of a conserved cognitive architecture.

We study the brains of "simple" animals

Studying microscopic animals, like the tiny worm C. elegans, hase yielded deep insights in molecular and cellular biology. We suspect the same will be true for neuroscience.

C. elegans has a mere 302 neurons, compared to human's 100 billiion neurons and the mouse's 1 billion. Yet this worm has a rich behavioral repertoire, capable of incredible feats of food finding, exploration, mating, toxin avoidance, predator escape, long-range migration, mating, social communication, and learning.

How does this organism execute these precision tasks and juggle them in a productive way? 

We also study tardigrades, affectionately known as water bears, which have 8 legs and can navigate rich microscopic three-dimensional environments, using a combination of coordinated gaits and independent limb control.

We study in vitro and in silico neural networks

About 500 million years ago, some organisms acquired the ability to grow  large pools of largely undifferentiated neurons and shape them through development and learning in order to flexibly solve problems and drive complex tasks, thereby getting around the limited information capacity of the genome. We study stripped-down versions of these "high-n" neural systems to understand how they perform rich cognitive tasks, and try to determine what ingredients, or rules of assembly and operation, are required in order for such sophisticated problem-solving functions to emerge. Our objects of study include cultured human neurons in controlled experimental conditions (organoids) and purely in silico neural networks.

We develop bleeding-edge methods to acquire and analyze data.

We live in a revolutionary time where both computational power and data acquisition power are exploding. We combine high-speed volumetric microscopy, electrical engineering, microfluidics, genomic engineering, machine learning, and modern software paradigms to build experimental systems of unprecedented throughput and interrogative control.


With big data comes the challenge of data interpretation. We develop novel computational methods to analyze data, facilitate discovery, and generate insight.  

bottom of page