Contextuality and inductive bias in quatum machine learning models: numerical study

This repository contains the code (numerics.ipynb) used to generate the plots of section 8 'outperforming classical surrogates' in arXiv:XXXXX.

The quantum circuit simulation is done in pennylane. Both the quantum and surrogate training using the JAX for vectorisation and gradient computation.

Dependencies:

• pennylane
• JAX
• optax

The files X_data.npy and Y_data.npy contain the input data and labels in numpy array format.

please email any questions to bowles.physics@gmail.com.