RT Journal Article SR Electronic T1 Unsupervised discovery of demixed, low-dimensional neural dynamics across multiple timescales through tensor components analysis JF bioRxiv FD Cold Spring Harbor Laboratory SP 211128 DO 10.1101/211128 A1 Alex H. Williams A1 Tony Hyun Kim A1 Forea Wang A1 Saurabh Vyas A1 Stephen I. Ryu A1 Krishna V. Shenoy A1 Mark Schnitzer A1 Tamara G. Kolda A1 Surya Ganguli YR 2017 UL http://biorxiv.org/content/early/2017/10/30/211128.abstract AB Perceptions, thoughts and actions unfold over millisecond timescales, while learned behaviors can require many days to mature. While recent experimental advances enable large-scale and long-term neural recordings with high temporal fidelity, it remains a formidable challenge to extract unbiased and interpretable descriptions of how rapid single-trial circuit dynamics change slowly over many trials to mediate learning. We demonstrate a simple tensor components analysis (TCA) can meet this challenge by extracting three interconnected low dimensional descriptions of neural data: neuron factors, reflecting cell assemblies; temporal factors, reflecting rapid circuit dynamics mediating perceptions, thoughts, and actions within each trial; and trial factors, describing both long-term learning and trial-to-trial changes in cognitive state. We demonstrate the broad applicability of TCA by revealing insights into diverse datasets derived from artificial neural networks, large-scale calcium imaging of rodent prefrontal cortex during maze navigation, and multielectrode recordings of macaque motor cortex during brain machine interface learning.