SignalGP-Lite: Event Driven Genetic Programming Library for Large-Scale Artificial Life Applications

Event-driven genetic programming representations have been shown to\noutperform traditional imperative representations on interaction-intensive\nproblems. The event-driven approach organizes genome content into modules that\nare triggered in response to environmental signals, simplifying simulation\ndesign and implementation. Existing work developing event-driven genetic\nprogramming methodology has largely used the SignalGP library, which caters to\ntraditional program synthesis applications. The SignalGP-Lite library enables\nlarger-scale artificial life experiments with streamlined agents by reducing\ncontrol flow overhead and trading run-time flexibility for better performance\ndue to compile-time configuration. Here, we report benchmarking experiments\nthat show an 8x to 30x speedup. We also report solution quality equivalent to\nSignalGP on two benchmark problems originally developed to test the ability of\nevolved programs to respond to a large number of signals and to modulate signal\nresponse based on context.\n