Assumption: Fixed-shot statistical model

A positive integer $n$ is chosen before data collection. The $n$ source preparations satisfy Assumption ass:source; the one-shot device is repeated with independent fresh environments and without access to the hidden source labels. The reduced input on the data registers is ${{\rho }_{\theta }}^{\otimes n}$, and the ideal repeated record map is ${\mathcal{I}}^{\otimes n}$. There is no unaccounted adaptive stopping, record deletion, or selection of successful trials.