Speaker
Description
Current and next-generation gravitational wave detectors are designed by human experts who must balance many coupled physical effects across multiple domains. At the same time, AI-based design methods are increasingly capable of proposing new experimental layouts and discovering novel measurement schemes, offering a complementary design paradigm with biases distinct from those of human designers. Enabled by Differometor, a differentiable frequency-domain interferometer simulator built for high-performance optimization, we cast the discovery of novel gravitational wave measurement techniques as an optimization problem over a vast space of hardware configurations. We introduce StrainForge, a growing benchmark for computational discovery in gravitational wave detector design. StrainForge provides a challenging testbed for exploring how optimization algorithms perform in the complex, high-dimensional design space of interferometric detectors. We present baseline results from a diverse set of optimization methods and invite suggestions for interesting design problems that can extend the benchmark.