Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
This paper introduces Lifting, a novel quantum state discrimination (QSD) framework designed to improve the reliability of mid-circuit measurements (MCMs) on noisy superconducting quantum hardware. Lifting addresses limitations in QSD accuracy and state deterioration by employing repetition encoding of joint in-phase and quadrature (I-Q) readout signals and contextual neural decoding with in situ calibration. Experimental results on IBM processors demonstrate the framework's effectiveness across various quantum workloads, suggesting a path towards more robust quantum error correction and control.
Quantum state discrimination errors are reduced by 2x by encoding auxiliary qubit readout signals and dynamically calibrating neural decoders to the evolving circuit environment.
Quantum computing promises new capabilities beyond classical computation. Mid-circuit measurements (MCMs) enable error correction and outcome-dependent control, but their reliability is limited by noise on superconducting hardware. These operations depend on accurate quantum state discrimination (QSD), which assigns measured signals to state labels. Neural network-based (NN-based) QSD improves this assignment, yet its error remains bounded below by the Bayes risk of the available readout distributions. A complementary challenge is state deterioration as successive measurements accumulate and circuit execution advances, making static calibration less representative of the workload. We present Lifting, a QSD framework that addresses both challenges through repetition encoding, neural decoding, and in situ calibration. Encoding supplies joint in-phase and quadrature (I-Q) readout signals from the target and auxiliary qubits, while contextual calibration trains the decoder for the conditions encountered during execution. Our theoretical analysis shows that retaining additional channels from a fixed joint distribution cannot increase Bayes risk and establishes sufficient conditions for contraction of an overlap-based upper bound. Experiments across three IBM processors evaluate the framework on repeated stabilizer measurement, parity-check, and teleportation workloads.