What problem does it solve? Implementing a Continuous Variable Quantum Neural Network (CVQNN) requires deep knowledge of CV quantum optics, truncated Fock space simulation, and gradient-based training of quantum circuit parameters. This Skill provides a complete, documented implementation of CVQNN for binary classification, covering encoding, variational layers, training, evaluation, and visualization. ## Core Features & Use Cases - End-to-End CVQNN Training: Encodes 2D features as coherent-state displacements, applies variational layers (beamsplitter, Kerr, displacement, squeezing, rotation), and trains with PyTorch Adam and MSE loss. - Fock Space Simulation: Builds CV gate unitaries (displacement, squeezing, rotation, Kerr) via torch.matrix_exp over complex128 matrices with configurable cutoff dimension. - Metrics and Visualization: Produces training loss curves, decision boundary plots, and circuit topology SVGs, returning accuracy, loss, and timing in a structured result dictionary. - Use Case: A quantum machine learning researcher wants to classify a two-moons dataset using a photonic quantum model. The Skill runs CVQNNAlgorithm with configurable layers, cutoff, epochs, and learning rate, returning final accuracy and saved plots. ## Quick Start Ask the AI to train a CVQNN classifier on a two-feature binary dataset such as sklearn's make_moons with 2 layers, cutoff 6, and 40 epochs, then report the final accuracy and loss.
