Evolutionary Game Theory Reveals Modular Specialization as a Universal Solution to Neural Trade-offs

This dataset contains the simulation results, source code, and analysis of an evolutionary game theory model investigating neural architectural trade-offs. We prove that modular neural architectures (segregated transport and memory) emerge as evolutionarily stable strategies (ESS) when cognitive systems face conflicting computational demands (high-noise transport vs. low-noise memory). Using replicator dynamics with 90 competing strategies across 441 environmental conditions, the data demonstrates that modular specialization yields a +6% to +31% reproductive fitness advantage over global architectures, leading to a 99.86% population takeover.