Spin network

A spin network is a graph whose edges are labeled by irreducible representations of a Lie group (typically SU(2)) and whose vertices are labeled by intertwiners — invariant tensors that describe how the representations at incident edges combine. Introduced by Roger Penrose in 1971 as a combinatorial model of quantum geometry, spin networks were later adopted as the fundamental kinematical states of loop quantum gravity, where they represent quantized three-geometries.

In loop quantum gravity, a spin network state describes a quantum geometry in which area and volume are discrete, quantized in units of the Planck scale. The edges of the network carry quanta of area; the vertices carry quanta of volume. The theory predicts that the geometry of space at the smallest scales is not a smooth manifold but a superposition of spin network configurations, with dynamics governed by the action of constraints on the space of these networks.

Spin networks are structurally related to tensor networks, which appear in condensed matter physics and quantum information theory. Both encode quantum states as networks of local tensors, but spin networks carry additional geometric interpretation: the labels on edges correspond to quantum numbers of angular momentum, and the network as a whole represents a state of quantum geometry rather than a generic quantum state. The relationship between spin networks and tensor networks remains an active area of research, with proposals that tensor network renormalization may describe the continuum limit of spin network dynamics.