A novel statistical algorithm for mining high-dimensional spike train (count) data for significant spatio-temporal patterns reveals new insights into task and brain area dependent functional organization of neural activity.

Independent coding without synaptic coordination explains complex sequences of population activity observed during theta states and maximizes the number of distinct environments that can be encoded through population theta sequences.

Phase-locking of hippocampal theta and gamma waves has been proposed to support memory formation, but an analysis using robust statistical methods finds no convincing evidence for the phenomenon.

A computational framework and analysis of rodent CA1 data reveals that theta oscillations sequentially sweep through the past, present, and future at a spatial scale that varies with the characteristic speed of the animal.

A combination of signal processing and machine learning form a new approach to classify oscillatory coupling in single cycles without averaging over time and to capture cycle-by-cycle changes in coupling.

Building on simple unsupervised matrix factorization techniques, the seqNMF algorithm successfully recovers neural sequences in a wide range of simulated and real datasets.

Techniques are presented to facilitate widespread and standardized chronic use of Neuropixels probes for high-yield, long-term neural recording in freely moving animals.

Interrogating mouse locus coeruleus via large-scale multipatch whole-cell recordings uncovers its cellular heterogeneity and cell-type-specific wiring logic.

A spatially-tuned normalization model accounts for neuronal responses to attended or unattended stimuli that are presented inside the classical receptive field or the surround, and explains various other observations.

A reusable system of 3D printed stainless steel microdrive and plastic head cap allows stable recording of electrophysiological data in mice and rats and easy recovery of silicon probes.