Breaking the Cascade: Compact Nonlinear Optical Computing with Single-Layer Encoder-Decoder Co-Localization

2026-05-31

AI Summary

This work addresses the challenge of achieving compact and efficient nonlinear optical computing without relying on nonlinear optical materials or multilayer diffractive structures. The authors propose a single-layer phase-only diffractive planar architecture that co-locates a dynamic encoder with a static, optimized decoder, enabling programmable nonlinear mappings through free-space propagation and intensity detection under coherent illumination. They demonstrate for the first time that a single diffractive surface can act as a universal approximator capable of accurately implementing arbitrary real-valued band-limited nonlinear functions. By introducing a post-training frozen phase bias, the system achieves enhanced expressive power and robustness against phase quantization errors. Experimentally, the platform simultaneously approximates nine distinct nonlinear functions—including common neural network activation functions and complex-valued mappings—in a single forward pass at visible wavelengths.

Abstract

We demonstrate that nonlinear computing can be achieved with a single linear diffractive surface under coherent illumination. We introduce a compact encoder-decoder co-localization (E+D) architecture in which an input-dependent dynamic encoder and a static optimized decoder are integrated within the same phase-only diffractive plane. Following free-space propagation, coherent interference between the encoder and decoder fields, combined with intensity detection, generates programmable nonlinear input-output mappings without requiring nonlinear optical materials or multiple diffractive layers. We prove that the proposed E+D optical processor is a universal approximator for arbitrary real-valued band-limited nonlinear functions and identify the physical factors governing its approximation fidelity, including the decoder degrees-of-freedom, detector aperture, and axial propagation distance. Crucially, we demonstrate that introducing a trained, frozen phase bias to the encoder region systematically enhances functional expressivity, providing robustness against coarse phase quantization on spatial light modulators. Using this framework, we accurately synthesize diverse nonlinear functions, including commonly used neural network activation functions and complex-valued nonlinear functions. Finally, we experimentally validate the proposed approach using a visible-light optical set-up trained through in situ learning, demonstrating the parallel approximation of 9 nonlinear functions in a single optical forward pass. By collapsing nonlinear optical computation into a single diffractive surface, the E+D architecture substantially reduces hardware and alignment complexity while preserving powerful function-approximation capabilities, providing a compact and scalable framework for analog information processing.

Problem

• nonlinear optical computing
• diffractive optics
• function approximation
• analog information processing
• encoder-decoder co-localization

Innovation

• nonlinear optical computing
• diffractive optics
• encoder-decoder co-localization
• universal approximation
• phase-only modulation