AI Summary
This study by Manley et al. explores cortex-wide neuronal dynamics using recordings from up to 1 million neurons. The researchers find that the dimensionality of these neuronal populations increases as more neurons are recorded. The most reliable dimensions are distributed across the cortex, separate from noise, and do not correlate with sensorimotor variables. This suggests that the complexity of neural dynamics scales with the number of neurons recorded.
Neuronal populations have often been argued to exhibit low-dimensional dynamics. Manley et al. record cortex-wide neuronal dynamics and demonstrate that the observed dimensionality is ever increasing with the number of simultaneously recorded neurons. Most reliable dimensions were distributed cortex wide, distinct from noise, yet exhibited no correlation with observed sensorimotor variables.