Fig. 2: Two-dimensional GPs. | Nature Physics

Fig. 2: Two-dimensional GPs.

From: Quantum neural networks form Gaussian processes

Fig. 2: Two-dimensional GPs.The alternative text for this image may have been generated using AI.

We plot the joint probability density function, as well as its scaled marginals, for the measurement outcomes at the output of a unitary Haar-random QNN acting on n = 18 qubits. The measured observable is Oj = Z1, where Z1 denotes the Pauli z operator on the first qubit. Moreover, the input states are for the left column, ρ1 = |0〉 〈0|n and ρ2 = |GHZ〉 〈GHZ| with \(\left\vert {\rm{GHZ}}\right\rangle =\frac{1}{\sqrt{2}}({\left\vert 0\right\rangle }^{\otimes n}+{\left\vert 1\right\rangle }^{\otimes n})\), and for the right column, ρ1 and ρ3 = |Ψ〉 〈Ψ| with \(\left\vert \varPsi \right\rangle =\frac{1}{\sqrt{d}}{\left\vert 0\right\rangle }^{\otimes n}+\sqrt{1-\frac{1}{d}}{\left\vert 1\right\rangle }^{\otimes n}\). In both cases we took 104 samples.

Source data

Back to article page