What does it do:

Explore, optimize, and benchmark circuits and estimators for quantum sensing protocols. The quantum probe is represented as parameterized quantum circuits, and the estimators as classical neural networks.

Basic usage:

import jax
import jax.numpy as jnp
from queso.sensors import Sensor
from queso.estimators import BayesianDNNEstimator

sensor = Sensor(n=4, k=4)

theta, phi, mu = sensor.theta, sensor.phi, sensor.mu
sensor.qfi(theta, phi)
sensor.cfi(theta, phi, mu)
sensor.state(theta, phi, mu)

data = sensor.sample(theta, phi, mu, n_shots=10)

estimator = BayesianDNNEstimator()
posterior = estimator(data)

Install

pip install git+https://github.com/benjimaclellan/queso.git

Quantum circuit simulations are done with tensorcircuit with JAX as the differentiable programming backend. Neural networks are also built on top of JAX using the flax library.

Citing

@article{maclellan2024endtoend,
      title={End-to-end variational quantum sensing}, 
      author={Benjamin MacLellan and Piotr Roztocki and Stefanie Czischek and Roger G. Melko},
      year={2024},
      eprint={2403.02394},
      archivePrefix={arXiv},
      primaryClass={quant-ph}
}

Acknowledgements

This project is supported by the Perimeter Institute Quantum Intelligence Lab and the Institute for Quantum Computing.