The geometry of cognition: from neural populations to neurotechnology

Abstract

Cognition is implemented by distributed neural populations whose collective activity evolves through structured, low-dimensional state spaces. In this review, we survey the analytical and conceptual tools used to study latent population dynamics in biological neural networks. We begin with single-unit coding, which remains essential for interpreting neuronal selectivity, and then review population-level approaches, including population vectors, cosine similarity, principal component analysis and its mnemonic variants, multivariable linear regression axes, demixed principal component analysis, and single-trial latent-variable models, and distinguish operationally defined population subspaces from neural manifolds with reproducible geometry or topology. We use these methods to discuss sensory coding, working memory, decision-making, temporal processing, and sensorimotor transformations. Two organizing principles are emphasized throughout. First, cognitive variables can be expressed in separable or near-orthogonal population dimensions, reducing interference between sensory, mnemonic, decision-related, and motor signals. Second, such geometric separation does not require anatomically or functionally discrete cell classes; separable population readouts can coexist with a continuum of mixed-selective single-neuron responses within a shared biological substrate. We then consider recurrent neural networks as testbeds for generating and evaluating mechanistic hypotheses about how circuits produce low-dimensional dynamics, and intracortical brain-computer interfaces as a translational domain in which population geometry becomes an operational control signal. Overall, this review argues that population dynamics now provide a common language for linking neurophysiology, computational modelling, and neurotechnology.

Abstract: arXiv descriptive metadata (CC0). Thank you to arXiv for use of its open access interoperability. This service was not reviewed or approved by, nor does it necessarily express or reflect the policies or opinions of, arXiv.