arXiv subjects

Vivian Maloney

Publications and source records attributed to Vivian Maloney.

3 recordsLinked to original sources

arXiv2026

Revealing Noise in Axial Motion Through Quantum Noise Spectroscopy on a Trapped Ion Processor

Native noise processes in quantum processors are often difficult to isolate because multiple error mechanisms contribute to the same measured loss of coherence. Here we use dephasing-robust quantum noise spectroscopy to identify and characterize control noise induced by axial motion in an individually-addressed trapped-ion processor. When the ion motion is transverse to the addressing beam, thermal axial motion couples to the beam profile and produces effective amplitude control noise. We show that this noise is governed primarily by the local beam curvature and appears as a low-frequency contribution to the reconstructed control-noise spectrum. By varying the ion position within the beam profile, we separate curvature-dependent axial-motion noise from curvature-independent native control noise and extract motional parameters that are otherwise difficult to access on this platform. We also apply the protocol in parallel to a four-ion register, demonstrating a spectroscopic method for simultaneous characterization of position-dependent control noise across multiple qubits. The results of this study identify beam inflection points as operating regions that suppress axial-motion-induced noise at the cost of reduced Rabi rate, as found in PRX Quantum 3, 010334 (2022).

quant-ph↗

arXiv2024

High-Resolution Convolutional Neural Networks on Homomorphically Encrypted Data via Sharding Ciphertexts

Recently, Deep Convolutional Neural Networks (DCNNs) including the ResNet-20 architecture have been privately evaluated on encrypted, low-resolution data with the Residue-Number-System Cheon-Kim-Kim-Song (RNS-CKKS) homomorphic encryption scheme. We extend methods for evaluating DCNNs on images with larger dimensions and many channels, beyond what can be stored in single ciphertexts. Additionally, we simplify and improve the efficiency of the recently introduced multiplexed image format, demonstrating that homomorphic evaluation can work with standard, row-major matrix packing and results in encrypted inference time speedups by $4.6-6.5\times$. We also show how existing DCNN models can be regularized during the training process to further improve efficiency and accuracy. These techniques are applied to homomorphically evaluate a DCNN with high accuracy on the high-resolution ImageNet dataset, achieving $80.2\%$ top-1 accuracy. We also achieve an accuracy of homomorphically evaluated CNNs on the CIFAR-10 dataset of $98.3\%$.

cs.CR↗

arXiv2023

Quantum Control Noise Spectroscopy with Optimal Suppression of Dephasing

We extend quantum noise spectroscopy (QNS) of amplitude control noise to settings where dephasing noise or detuning errors make significant contributions to qubit dynamics. Previous approaches to characterize amplitude noise are limited by their vulnerability to low-frequency dephasing noise and static detuning errors, which can overwhelm the target control noise signal and introduce bias into estimates of the amplitude noise spectrum. To overcome this problem, we leverage optimal control to identify a family of amplitude control waveforms that optimally suppress low-frequency dephasing noise and detuning errors, while maintaining the spectral concentration in the amplitude filter essential for spectral estimation. The waveforms found via numerical optimization have surprisingly simple analytic forms, consisting of oscillating sine waves obeying particular amplitude and frequency constraints. In numerically simulated QNS experiments, these waveforms demonstrate superior robustness, enabling accurate estimation of the amplitude noise spectrum in regimes where existing approaches are biased by low-frequency dephasing noise and detuning errors.

quant-ph↗