Suppressing photon detection errors in LOQC

This project aims to simulate and optimize the six mode KLM controlled sign flip gate to achieve higher output fidelity for preparing the state $$\frac{1}{2}(\ket{0, 1, 0, 1} + \ket{1, 0, 0, 1}) + \ket{0, 1, 1, 0} - \ket{1, 0, 1, 0}$$ from the dual-rail encoded $\ket{+}\ket{+}$ states with imperfect detectors

Training and generating plots

Makes sure to install the dependencies using the following command:

pip install -r requirements.txt

You can execute the optimizations by running the trainer.py script.

Example

python trainer.py --output-dir ./results --starting-s-star 0.0650 --ending-s-star 0.0850 --enhanced --use-jit

This script will save the generated losses and weights in the specified output directory.

The plots can be generated using the script make-plots.py.

Example

python make_plots.py --input-dir ./results