Field-level perturbational complexity tracks neuronal reproducible dimensionality within brains, not across them

Abstract

Perturbational complexity is measured from field potentials, yet the quantity of interest lives in neuronal populations. Whether the field-level measurement reflects the reproducible richness of the underlying spiking activity - or a field-specific shadow - had never been tested, because it requires both signals in the same brain during the same perturbations across arousal states. Using the open Allen Institute dataset (DANDI 000458: simultaneous 30-channel epidural EEG and hundreds of Neuropixels single units in head-fixed mice during cortical electrical stimulation, awake and under isoflurane), this preprint computes debiased reproducible dimensionality (R-dim) from sorted spiking populations on exactly the matched trial sets of the author's published EEG analysis, under a preregistration sealed before any spike time was read. Results (n = 15 animals): levels do not correspond across animals (Spearman +0.25, n.s.) - both level quantities scale with their own, independent coverage; the state contrast has a genuine neuronal counterpart (spiking R-dim 8.48 awake vs 4.69 under isoflurane, p = 0.0027), spiking dimensionality runs about 3x the simultaneous field value, and within-animal state effects couple across levels (+0.64, p = 0.0054). The pattern survives both preregistered bin sizes. The minority of animals in which anesthesia increased field dimensionality showed the same reversal in their neurons (3/3); labeled exploratory analyses show spontaneous firing was suppressed uniformly across animals while evoked-response attenuation varied 1.1-7.9x and the state effect followed that attenuation: where the anesthetic failed to suppress responsiveness, dimensionality persisted. Practical conclusion: field-level complexity should not be compared across recordings, but it is a faithful within-brain tracker of neuronal dimensionality change - precisely the clinically relevant use.