Speaker
Description
We present fast algorithms to test unitary channels -- under the depolarizing noise assumption, often made in fault-tolerant quantum computations. Our optimal algorithms answer the following question: is the noisy unitary channel implemented by a given black box identical to a target unitary or $\varepsilon$-far from it in the diamond distance, assuming that the deviation is a depolarizing channel with unknown parameter? Our algorithm has a query complexity of $\Theta(1/\varepsilon).$ The query complexity of the relaxed problem of testing whether the black-box channel is $\varepsilon_1$-close to a target unitary or $\varepsilon_2$-far in the diamond distance is $\Theta\bigl(\varepsilon_2/(\varepsilon_2 - \varepsilon_1)^2\bigr).$ In both cases, we provide matching lower bounds that hold even for adaptive, ancilla-assisted protocols with multi-outcome incoherent measurements.
| Topic | Theoretical foundations of quantum learning |
|---|---|
| Additional Topic | Quantum state and process tomography with learning-based approaches |