Analysis of Elzerman readout based on Bayesian inference methods

Methods based on spin-to-charge conversion are commonly used to initialize and measure Loss-DiVincenzo qubits [1]. In this scheme, we tune the spin-up and spin-down energy levels of the dot relative to the Fermi energy of an electron reservoir, and deduce the spin of the electron on the dot based on whether it tunnels off the dot into the Fermi sea [2]. The presence or absence of a tunneling event is detected using a charge detector. Traditionally, the presence of tunneling has been defined by a constant threshold [3]. More sophisticated methods could both improve the accuracy and speed of detection of these events. In this talk, we report on the implementation of an algorithm based on Bayesian inference for the analysis of Elzerman readout data, and characterize its performance as compared to previous methods.