Quantum and optical technology

Abstract:

Photon echoes are generated on the 6S1/2-6P1/2 transition in Cs vapor with excitation pulses short enough to excite all hyperfine states. As the excitation pulse separation is varied, the temporal profile of the echo reshapes and the time-integrated echo intensity modulates. These effects are typical of all the alkali vapors and provide the simplest, nontrivial display of quantum beating in a four-level system. The present paper describes a detailed experimental study of both kinds of modulation effects and utilizes the billiard ball model to analyze them. Copyright © 1996 by MAHK Hayκa/Interperiodica Publishing.

Abstract:

Backpropagation through nonlinear neurons is an outstanding challenge to the field of optical neural networks and the major conceptual barrier to all-optical training schemes. Each neuron is required to exhibit a directionally dependent response to propagating optical signals, with the backwards response conditioned on the forward signal, which is highly non-trivial to implement optically. We propose a practical and surprisingly simple solution that uses saturable absorption to provide the network nonlinearity. We find that the backward propagating gradients required to train the network can be approximated in a pump-probe scheme that requires only passive optical elements. Simulations show that, with readily obtainable optical depths, our approach can achieve equivalent performance to state-of-the-art computational networks on image classification benchmarks, even in deep networks with multiple sequential gradient approximations. This scheme is compatible with leading optical neural network proposals and therefore provides a feasible path towards end-to-end optical training.