Given the predictions and truth values, return a number between 0 and 1 indicating the accuracy of predictions.

Using get_probs() above, take only the probability of measuring the bitstring of all zeroes (i.e, take the projector |0⟩^(⊗n)⟨0|^(⊗n)) on the

Randomly generate initial parameters. We need initial parameters for the variational circuit ansatz implementing W(alpha) and the standard deviati

Computes the coefficients for the classical model, E^(A), E^(B), E^(C), and E^(D).

For a given threshold, get the predicted labels (1 or -1), given the anomaly scores.

Check whether an operator is orthogonal to a space given by some basis.

Ising transition line

Kosterlitz-Thouless transition line

This function contains all gates that precede parametrized layer 0

This function contains all gates that precede parametrized layer 1

Measure all X type stabilizers based on the parity check matrix X then apply Z type corrections from our decoder :param H: Parity check X mat

Probability of measuring |0…0⟩ after U(x) U†(y).

A bias-invariant unitary (U in the paper)

Helper function to plot vectors on a sphere.

Process a block of 'text'. Return a markdown text string representing the content.

Ansatz of the QCNN model Repetitions of the convolutional and pooling blocks until only 2 wires are left unmeasured

One round of Steane code QEC with LUT decoding.

Creates the recurrent loop for the Recurrent Neural Network.

Convert a state vector into {bitstring: amplitude | probability}.

Sum a list of pytrees.

bias-invariant unitary with swap matrix as generator

n (int): number of variables time_single_case (float): time to solve a single case

Computes B^T y into the syndrome register.

Controlled versions of the unitary components A_l of the problem matrix A.

Apply CNOTs in a ring pattern

Take the classical expectation value of of the projector on zero sampling the parameters of D from normal distributions. The expectation value is

Initializes and returns a custom Keras model which is ready to be trained.

Performs the quantum Chebyshev transform.

Layer of parametrized qubit rotations around the y axis.

Implements the Split & Cycle shift unitary.

Unitary matrix rotating the ground state to the problem vector |b> = U_b |0>.

Circuit implementing the basis rotations of the Hamiltonian fragments.

Args: csv_file (string): Path to the csv file with annotations. root_dir (string): Directory with all the images.

Format author information as a dictionary.

Pre‑compute variable‑node and check‑node neighbors. Returns ------- var_neighbors : tuple[tuple[int, ...], ...] # length = n ch

Load all of the iterator into a list.

Load an inventory from a local file. Returns None if file missing or invalid.

Download objects.inv from URL and load with retry logic. Returns (inventory, raw_bytes).

Generate datapoints arranged in an even circle.

Build a name -> item dict for easy lookup of inventory items.

Adapts the shape of an array to match the reference (either a batched vector or matrix). Example: >>> a = jnp.ones(3) >>> b = jnp.ones((3,

Apply the unitary matrix of the gate to a batch of states.