Audrey Sederberg

The paper analyzes spatiotemporal dynamics of spontaneous activity in developing ferret visual cortex, finding repeated, stereotyped patterns and predictive spatial activity motifs.

The paper uses a hidden Markov model to relate latent dynamics of neural spiking activity in the somatosensory cortex to classical brain state metrics like local field potential fluctuations and behavior.

This paper develops real-time cortical state estimation algorithms using Gaussian Mixture Models and Hidden semi-Markov Models, showing they can robustly infer cortical states from electrophysiological data.

The paper shows that randomly connected neural networks can produce selectivity and decoding properties similar to real neural populations, suggesting emergent properties from random connectivity.

The paper improves sensory input detection in awake mice by using a cortical state classifier to account for trial-to-trial variability in neural responses.