Quantum Convolutional Neural Network: Domain-Wall Classifier (2D)

A quantum convolutional neural network compresses an 8-qubit spin chain down to a single classifying qubit through alternating layers of parameterized entangling convolution and measurement-based pooling — the same architecture, due to Cong, Choi & Lukin, used to detect symmetry-protected topological phases on real quantum hardware. This simulator runs the genuine 256-amplitude statevector: pick an input pattern with zero, one or two domain walls, tune the convolution and pooling angles, and step through the three layers while watching each qubit's true Bloch vector — traced out of the full multi-qubit state — shrink from a ring of eight down to one qubit whose measurement probability is the network's answer. The ring diagram is a drag-to-rotate pseudo-3D projection rendered entirely on a 2D canvas, paired with a live purity sparkline and a domain-wall probability gauge.