Interpretable Geometry Sensitivity for Inverse Design of Integrated Photonics

TL;DR

This paper introduces an interpretability workflow for inverse photonic design that uses sensitivity maps to identify physically meaningful substructures, aiding fabrication and design-rule checking.

Contribution

It presents a novel pixel-level sensitivity analysis method using Integrated Gradients for inverse-designed photonic devices, enhancing interpretability and fabrication robustness.

Findings

High-attribution hotspots align with key physical substructures.

Controlled perturbations in sensitive regions significantly affect device performance.

The method improves understanding of inverse design layouts without modifying electromagnetic simulations.

Abstract

As an increasingly powerful technique in integrated photonics, inverse design uses optimization algorithms to automatically create compact, high-performance photonic structures, often yielding non-intuitive layouts far more compact than conventional designs. While adjoint-based inverse design is a prominent optimization method, the resulting free-form layouts are difficult to interpret or diagnose under fabrication variability, even for experienced photonic device designers. We present an experimentally validated interpretability workflow that produces pixel-level sensitivity maps directly on the binary mask of an inverse-designed device. Using wavelength-division demultiplexers (WDMs) at 1310/1550 nm as examples, we train a lightweight convolutional surrogate to regress figures of merit (FoMs) and apply Integrated Gradients (IG) to attribute predicted transmission to individual pixels. We demonstrate that high-attribution hotspots correspond to physically meaningful substructures, such as splitter hubs and high-curvature edges. Experimental results show that controlled perturbations in these high-sensitivity regions result in up to an 11x higher excess insertion loss compared to perturbations in non-sensitive regions, consistent with full-wave simulations. This approach adds a practical explainability layer to existing pipelines, offering a clear pathway for foundry-compatible design-rule checking and fabrication-aware constraint allocation without modifying the underlying electromagnetic solver.