Simone Fasciati

Abstract:

Abstract Decoherence in superconducting quantum circuits, caused by loss mechanisms like material imperfections and two-level system (TLS) defects, remains a major obstacle to improving the performance of quantum devices. In this work, we present atomic-level characterization of cross-sections of a Josephson junction and a spiral resonator to assess the quality of critical interfaces. Employing scanning transmission electron microscopy (STEM) combined with energy-dispersive X-ray spectroscopy (EDS) and electron-energy loss spectroscopy (EELS), we identify structural imperfections associated with oxide layer formation and carbon-based contamination, and correlate these imperfections to the pattering and etching steps in the fabrication process and environmental exposure. These results help to understand that TLS imperfections at critical interfaces play a key role in limiting device performance, emphasizing the need for an improved fabrication process.

Abstract:

A qutrit represents a three-level quantum system, so that one qutrit can encode more information than a qubit, which corresponds to a two-level quantum system. This work investigates the potential of qutrit circuits in machine learning classification applications. We propose and evaluate different data-encoding schemes for qutrits, and find that the classification accuracy varies significantly depending on the used encoding. We therefore propose a training method for encoding optimization that allows to consistently achieve high classification accuracy, and show that it can also improve the performance within a data re-uploading approach. Our theoretical analysis and numerical simulations indicate that the qutrit classifier can achieve high classification accuracy using fewer components than a comparable qubit system. We showcase the qutrit classification using the encoding optimization method on a superconducting transmon qutrit, demonstrating the practicality of the proposed method on noisy hardware. Our work demonstrates high-precision ternary classification using fewer circuit elements, establishing qutrit quantum circuits as a viable and efficient tool for quantum machine learning applications.