Skip to main content
. 2025 Jul 7;12(8):nwaf269. doi: 10.1093/nsr/nwaf269

Figure 3.

Figure 3.

Schematic of NN-based QST. (a) Obtain the measured frequencies Inline graphic and the ground truth of Inline graphic via Cholesky decomposition for state Inline graphic. (b) A multilayer NN model to map the frequency vector Inline graphic to a predicted real vector Inline graphic, and the NN model is trained to minimize the mean squared error between the predicted Inline graphic and the expected value Inline graphic. (c) Obtain a physical density matrix Inline graphic from the optimal Inline graphic. The notation Arg indicates the complex phase (in radians) of each density matrix element.