Quantum Circuit Born Machines

Quantum Circuit Born Machines

This PennyLane demo implements the Quantum Circuit Born Machine (QCBM), a generative model that uses a parameterized quantum circuit to learn and sample from a target probability distribution. The model takes its name from the Born rule of quantum mechanics: the probability of measuring a given bitstring is the squared amplitude of the circuit's output state, so a parameterized circuit naturally defines a probability distribution over bitstrings, and training the circuit shapes that distribution to match data. The tutorial builds a QCBM in PennyLane, choosing an expressive ansatz, and trains it to reproduce a target distribution by minimizing a distance measure between the circuit's output distribution and the target, such as the maximum mean discrepancy, using gradient-based optimization. It walks through defining the model and loss, optimizing the parameters, and comparing the learned distribution against the target, and discusses why gradient estimation for distribution-matching losses requires care. As a purely generative, sampling-based model that needs no labels, the QCBM illustrates a distinct paradigm of quantum machine learning, generative modeling via the Born rule, presented hands-on in PennyLane.

Run it

pip install -r requirements.txt
python demo.py

Source and license

Imported from demonstrations_v2/tutorial_qcbm/demo.py in PennyLaneAI/demos at c52c0abeb5122218aa96b38eea848864cce7323f, under the Apache License 2.0. Original authors: Xanadu and the PennyLane community. The upstream LICENSE is included alongside this example.