Symmetry-preserved cost functions for variational quantum eigensolver

Hamzat Akande,Bruno Senjean,Matthieu Saubanere
2024-11-26
Abstract:Hybrid quantum-classical variational algorithms are considered ideal for noisy quantum comput- ers, as they significantly reduce quantum circuit depth compared to fully quantum methods like quantum phase estimation. This reduction requires a classical variational optimization task. Ide- ally, circuits are shallow enough to avoid quantum noise, and the cost function is convex but not flat, enabling efficient optimization. However, quantum circuits based on unitary coupled clus- ter ansätze scale quartically with the number of qubits. Hardware-Efficient Ansätze (HEA) offer shallower circuits but suffer from optimization issues, including symmetry breaking. We propose encoding symmetry preservation directly into the cost function, enabling more efficient use of HEA. Our variational, iterative algorithm controls circuit depth and optimization challenges.
Quantum Physics
What problem does this paper attempt to address?