Characterizing the Impact of Noise on Neural Computation

Explore the intricate relationship between noise and neural computation in this 33-minute lecture by Alex Williams from NYU and the Flatiron Institute. Delve into the ubiquitous nature of noise in biological neural circuits and its incorporation into artificial neural networks. Examine the challenges of accurately estimating noise statistics in neuroscience and discover a new statistical model that leverages experimental paradigm smoothness for efficient trial-to-trial noise covariance estimation. Learn about a novel measure of neural representational similarity for stochastic networks, designed to compare noise structure between artificial and biological systems. Gain insights into non-deterministic modes of neural network function and their potential implications for understanding lower-level intelligence from AI, psychology, and neuroscience perspectives.