Automated Discovery of Photonic Quantum Advantage Scaling Laws

Complete research compendium for "Automated Discovery of Photonic Quantum Advantage Scaling Laws" submitted to Physical Review Letters (2026). This repository contains all materials needed to reproduce the results: manuscript source files (LaTeX), publication-ready figures, datasets comprising 7,560 Gaussian Boson Sampling (GBS) simulations spanning experimentally accessible parameter space, analysis code, and comprehensive supplementary materials. Using Survivorship-Biased Conjecture Generation (SB-CG), we discover four quantitative scaling laws with exceptional stability (>0.99): universal photon number degradation F ∝ exp(-0.25N), cumulative loss F ∝ (1-L)^10.5D with the first precise coefficient determination, a novel N·η compensation mechanism showing detector efficiency partially offsets photon losses, and steep detector efficiency power law η^17.3. All discoveries are validated through three independent methods: synthetic data with known ground truth, consistency with published USTC Jiuzhang and Xanadu Borealis experimental parameters, and independent physics-based simulations achieving mean 17% agreement in the quantum advantage regime. This work provides the first quantitative resource equations for photonic quantum advantage, enabling hardware optimization and testable predictions.