Universal Quantum Error Mitigation via Random Inverse Depolarizing Approximation

Abstract

Given the severity of noise in near-term quantum computing, error mitigation is essential to reduce error in quantum-computer-generated expectation values. We introduce RIDA (Random Inverse Depolarizing Approximation), a simple universal method that harnesses randomly generated circuits to estimate a given circuit's global depolarization probability and corresponding error-free expectation value. Numerical tests indicate RIDA outperforms key benchmarks, suggestive of significant accuracy improvements for applications of quantum computing across fields including physics and chemistry.