In this “Brain Inspired” episode, Paul Middlebrooks is joined by Tatiana Engel, associate professor at the Princeton Neuroscience Institute. Engel builds latent circuit models—which explicitly model the hidden variables driving task performance—to explain how low-dimensional neural population activity is embedded in and emerges from the much-higher-dimensional connectivity structure of neural circuitry. They also discuss how the concept of manifolds will continue to inform neuroscience explanations, and Engel’s study with International Brain Laboratory measuring single-neuron firing patterns across the entire brain.
Read the transcript.
