AMP: Associative memory using glassy confocal cavity QED
Abstract
The goal of this project is to use the novel apparatus ARO has helped us to fund for the purpose of exploring quantum-optical neuromorphic optimization with an artificial spin glass. The apparatus is a one-of-a-kind cavity QED machine that employs a multimode cavity to couple atoms via intracavity photons. Our primary goal will be to use this experimental quantum-optical platform as an associative memory, a fundamental neural network capability that is a good starting point for exploring neuromorphic optimization in general. Our recent theory paper (Marsh et al., Physical Review X, 2021) showed that the system can serve as an Ising optimizer for associative memory by exploiting driven-dissipative quantum dynamics. Surprisingly, it supports greater memory capacity than extant associative memory methods. We now have the opportunity to study the first cavity QED device that, guided by our practicable theory roadmap, may realize neural-network-like optimization capabilities. This abstract is publicly releasable.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Sep 28, 2022
- Source ID
- W911NF2210261
Entities
People
- Benjamin Lev
Organizations
- Army Contracting Command
- Stanford University
- United States Army