Variational Quantum Classifier

This simulator trains a real single-qubit machine-learning model in your browser. Two-dimensional data points are encoded into a qubit's state with fixed rotation gates, a small set of trainable rotations reshapes that state, and the qubit's Z-expectation value becomes the model's prediction. Gradients are computed with the parameter-shift rule — the exact technique used to train variational circuits on real quantum hardware — and applied with gradient descent step by step. Watch the Bloch sphere fill with every data point's quantum state, the decision plane separate the two classes as loss falls, and the 2D feature-space panel trace the resulting boundary live.