OL-Transformer: A Fast and Universal Surrogate Simulator for Optical Multilayer Thin Film Structures

TL;DR

OL-Transformer is a universal, fast surrogate model that accurately predicts optical spectra for a vast array of multilayer thin film structures, significantly accelerating simulations while maintaining high accuracy.

Contribution

The paper introduces the OL-Transformer, a novel universal surrogate simulator capable of handling diverse structures using structure serialization and self-attention mechanisms.

Findings

Predicts spectra for up to 10^25 structures

Achieves six-fold faster simulation than physical solvers

Learns physical embeddings for accurate light-matter interaction modeling

Abstract

Deep learning-based methods have recently been established as fast and accurate surrogate simulators for optical multilayer thin film structures. However, existing methods only work for limited types of structures with different material arrangements, preventing their applications towards diverse and universal structures. Here, we propose the Opto-Layer (OL) Transformer to act as a universal surrogate simulator for enormous types of structures. Combined with the technique of structure serialization, our model can predict accurate reflection and transmission spectra for up to $10^{25}$ different multilayer structures, while still achieving a six-fold degradation in simulation time compared to physical solvers. Further investigation reveals that the general learning ability comes from the fact that our model first learns the physical embeddings and then uses the self-attention mechanism to capture the hidden relationship of light-matter interaction between each layer.