New Submission: Subspace corrections to quantum ansatzes

Due to high error rates from imperfect devices and the lack of expressivity of most quantum ansatze in hardware calculations, the practical usage of quantum devices remains limited. In the current pre-fault-tolerant hardware, error mitigation techniques are essential to improve computational accuracy in quantum hardware calculations. In this presentation, we introduce a subspace correction framework, based on classical equation-of-motion theory, designed to improve the molecular ground state energy achieved from quantum ansatze. The correction improves the computational results by (a) increasing the subspace of the solution, leading to improved expressivity of the ansatz, and (b) mitigating the effects of noise in the hardware. Additionally, we show promising examples and strategies to implement our idea cost-effectively on quantum hardware.