Physically motivated extrapolation for quantum error mitigation

Abstract: Quantum error mitigation techniques are essential for the current NISQ and emerging Megaquop-era machines, which, despite their noise, are capable of performing utility-scale quantum computations. However, most QEM methods incur exponential sampling overhead to achieve unbiased estimates, limiting their practical applicability. Recently, it was shown that by using error mitigation by restricted evolution (EMRE), expectation values of a physical observable can be obtained in constant sampling overhead at the cost of a non-zero bias, which grows as the circuit size or hardware noise increases. To overcome this problem, we introduce the Physics-Inspired Extrapolation (PIE) method, built upon the EMRE framework, to achieve enhanced accuracy and robustness. Unlike traditional zero-noise extrapolation, our method assigns operational interpretation to the parameters in the extrapolation function used in PIE. We demonstrate the efficacy of this method on IBMQ hardware and apply it to simulate 84-qubit quantum dynamics efficiently. Our technique yields accurate results with significantly smaller variance, establishing PIE as a practical and scalable error mitigation strategy for near-term and early fault-tolerant quantum computing.