The four N‑dimensions are not separate fields but one commuting diagram: physics and computation are the hardware and learning faces; dynamics and NeuroAI are what emerges when the diagram closes.
NeuroPhysics
Statistical physics of cortical circuits. Field potentials, scaling, non-resistive extracellular media.
NeuroComputation
Information-theoretic treatments of depth. Coarse-graining and renormalisation as computation.
NeuroDynamics
Sleep oscillations, spindles, K-complexes — the transient grammar of thalamocortical states.
NeuroAI
The encompassing surface. Bridges between biological computation and machine learning.
The framework
Nima Dehghani investigates how neural systems compute across scales — from cortical microcircuits to coarse-grained representations in deep networks. The N⁴ programme treats neurophysics, neurocomputation, and neurodynamics as a single object viewed from three faces, with NeuroAI as the surface on which they meet.
This page is a living document. Notes, derivations, and the running argument for depth-as-renormalisation will be written here over time; the Paper Maze holds the discrete results, the Idea Garden the concepts that thread them.