Pattern formation and reservoir computation in activator–inhibitor cellular automata

This record contains four datasets associated with the manuscript “Computational Dynamics of Turing Patterns: Information Processing and Complexity in Activator–Inhibitor Reservoirs”. Dataset 1, “Turing CA Spatiotemporal Outputs and Complexity Sweeps: Sigmoid Activation”, contains spatiotemporal outputs and derived complexity summaries from activator–inhibitor cellular automaton simulations using the continuous sigmoid update rule. This dataset underlies the sigmoid analyses in Figures 2–5. Dataset 2, “Turing CA Spatiotemporal Outputs and Complexity Sweeps: Logistic–Step Activation”, contains spatiotemporal outputs and derived complexity summaries from activator–inhibitor cellular automaton simulations using the logistic–step relaxation rule. This dataset underlies the logistic–step analyses in Figures 2–5. Dataset 3, “Reservoir Computing X-bit Memory Test Results for Activator–Inhibitor Cellular Automata”, contains the results of the X-bit memory-task experiments, including tuning sweeps, source-geometry comparisons, and radius–radius performance landscapes. This dataset underlies Figures 6 and 7. Dataset 4, “Reservoir Computing MNIST Results for Activator–Inhibitor Cellular Automata”, contains the outputs of the MNIST image-classification experiments, including repeated classification sweeps across sample sizes and contour analyses over activator/inhibitor radii. This dataset underlies Figures 8 and 9.