Searcharxiv⌕ Search

arXiv subjects

Noah A. Crum

Publications and source records attributed to Noah A. Crum.

3 recordsLinked to original sources
arXiv2026

High-flux sub-Poissonian twin fields generation from warm atomic vapor

We demonstrate the generation of sub-Poissonian twin fields via near-degenerate spontaneous four-wave mixing (SFWM) in warm $^{85}\mathrm{Rb}$ vapor at 795~nm. When seeded with a weak coherent field, the generated twin beams exhibit approximately $5.5~\mathrm{dB}$ of intensity-difference squeezing in free space and retain about $3~\mathrm{dB}$ after coupling into polarization-maintaining (PM) fibers. Under vacuum seeding, time-resolved photon-counting measurements yield Mandel parameters of $Q\approx-0.7$ for each individual field, demonstrating strong photon-number squeezing. To explain these observations, we develop a finite-resource saturation model in which occupation-dependent SFWM gain, arising from competition for a finite nonlinear gain resource, suppresses large photon-number fluctuations within an effective collective mode selected by the PM-fiber spatial projection, thereby producing the observed negative Mandel-$Q$ parameters. The temporal cross-correlation between the twin photons exhibits a distinctive flat-topped profile resulting from the interplay of multiple $χ^{(3)}$ processes in the atomic medium and is in excellent agreement with the theoretical model. Combining high photon flux, near-resonant operation, robust sub-Poissonian photon statistics, and fiber compatibility, this source provides a promising platform for scalable quantum-enhanced sensing and quantum information processing.

quant-ph↗
arXiv2025

Stochastic Security as a Performance Metric for Quantum-enhanced Generative AI

Motivated by applications of quantum computers in Gibbs sampling from continuous real-valued functions, we ask whether such algorithms can provide practical advantages for machine learning models trained on classical data and seek measures for quantifying such impacts. In this study, we focus on deep energy-based models (EBM), as they require continuous-domain Gibbs sampling both during training and inference. In lieu of fault-tolerant quantum computers that can execute quantum Gibbs sampling algorithms, we use the Monte Carlo simulation of diffusion processes as a classical alternative. More specifically, we investigate whether long-run persistent chain Monte Carlo simulation of Langevin dynamics improves the quality of the representations achieved by EBMs. We consider a scheme in which the Monte Carlo simulation of a diffusion, whose drift is given by the gradient of the energy function, is used to improve the adversarial robustness and calibration score of an independent classifier network. Our results show that increasing the computational budget of Gibbs sampling in persistent contrastive divergence improves both the calibration and adversarial robustness of the model, suggesting a prospective avenue of quantum advantage for generative AI using future large-scale quantum computers.

cs.LG↗