Accelerating quantum error mitigation on quantum computers by several orders of magnitude

(QEC), QEM does not require additional hardware resources, making it a practical tool for experiments on today’s quantum processors. In the long term, QEM may also help overcome some limitations and imperfections of error-correction-based approaches. However, QEM often comes with a substantial sampling overhead. This leads to long runtimes, during which device noise can drift and reduce the reliability of standard mitigation protocols. Approaches based on agnostic noise amplification are naturally resilient to such noise variations, but their runtime cost has remained a major practical obstacle. Here, we introduce virtual noise scaling, a post-processing framework that reduces this sampling overhead by three to five orders of magnitude. With this reduction, mitigation tasks that once seemed unrealistic, even for future devices, become challenging but achievable. We validate the method on previously reported experimental data, showing substantial gains in both mitigation efficiency and accuracy.

Bio:

B.Sc in EE technion. M.Sc in Physics Technion, Supervisor Prof. Mordechai Segev,  topics: Spatial soliton. PhD in Physics Technion,  Supervisor Prof. Nimrod Moiseyev, topic: non-adibatic effects near effective non hermitian degeneracies in quantum mechanics. Postdoc, Hebrew University, Supervisor Prof. Ronnie Kosloff, topic: microscopic Quantum heat engines. PI at the Hebrew University, Main interests: quantum dynamics, quantum computers, quantum noise, quantum thermodynamics.