# Nonlinear optical extreme learner via data reverberation with incoherent light

## TL;DR

This paper introduces an energy-efficient optical machine learning system using incoherent light to achieve nonlinear transformations without complex materials.

## Contribution

A novel optical extreme learner using data reverberation in incoherent light to enable nonlinear learning with low power.

## Key findings

- The optical learner outperforms linear digital networks in image classification and XOR benchmarks.
- The system matches the accuracy of fully nonlinear digital models while using significantly less power.
- The approach reduces complexity, cost, and energy consumption for scalable optical machine learning.

## Abstract

Artificial neural networks have revolutionized fields from computer vision to natural language processing, yet their growing energy and computational demands threaten future progress. Optical neural networks promise greater speed, bandwidth, and energy efficiency but suffer from weak optical nonlinearities. Here, we demonstrate a low-power, incoherent-light-compatible optical extreme learner that leverages “data nonlinearity” from optical pattern reverberations, eliminating reliance on intrinsic nonlinear materials. By encoding input data in the spatial polarization distribution of a tailored optical cavity and allowing light to pass through it multiple times, we achieve nonlinear transformations at extremely low optical power. Coupled with a simple trainable readout, our optical learner consistently outperforms linear digital networks in standard image classification tasks and XOR benchmarks, delivering accuracy matching fully nonlinear digital models. Our compact, energy-efficient approach substantially reduces complexity, cost, and energy consumption, paving the way for practical, scalable all-optical machine learning platforms.

Incoherent-light reverberations enable a nonlinear optical learner matching digital nets at a fraction of the power.