Refrigeration via purification through repeated measurements

Breaking the limits of purification: postselection enhances heat-bath algorithmic cooling

“…When those dynamics can be arbitrarily controlled, the system can always trivially be cooled; this is because adding an auxiliary system, performing an arbitrary joint interaction, then ignoring or measuring the arbitrary system allows one to enact an arbitrary quantum channel on a given quantum system to generate any desired state [63] and strongly violates the assumptions of HBAC. The HBAC can also be replaced by any other procedure that helps to serially cool a quantum system toward its ground state, including other measurement-based protocols [16][17][18][19][20][21][22][23][24][25][26][27][28][29][30]. Third, HBAC protocols have recently been shown to potentially benefit from the presence of environmental noise [12].…”

“…However, this naive picture is too simplistic, as it requires n ideal binary-outcome measurements to purify n qubits. Alternative measurement-based purification schemes exist that require many measurements or only converge toward the purest state [16][17][18][19][20][21][22][23][24][25][26][27][28], such as a recent idea that proposes to perform ?n measurements on a single auxiliary qubit to purify n qubits (with some finite success probability that decays exponentially with n) [29] and one that seeks to reduce the number of measurements by using techniques from reinforcement learning [30]. In comparison, our method can completely purify n qubits with a single binary measurement (and arbitrarily high success probability).…”