Summary
Can a learned model accelerate predictions for layered VCSEL structures?
Trains transformer models on simulated Fabry–Pérot, DBR and VCSEL configurations.
Predicts resonant energies and Q-factors on held-out configurations.
Related modelling approaches
Relation labels come from the method guide and are editorial. Unlabelled links are conceptual connections or comparison methods; the paper does not necessarily implement them.
Commentary
Cavity-property prediction does not establish accurate gain saturation, polarization dynamics or out-of-distribution performance.
A task-specific surrogate; compare it with physics-constrained scattering surrogates rather than a full laser simulator.
Interpretation and connections in this note, separate from the authors’ reported findings.