// Gallery

Pattern Gallery

Nine showcase patterns from the Gray-Scott parameter space, each rendered at 256×256 resolution after 15,000 time-steps. These are not generated images or ML outputs — they are deterministic solutions to coupled partial differential equations, rendered directly from the simulation data.

Each pattern is labeled with its (F, k) parameters and a descriptive name based on the visual morphology. The patterns span the full diversity of the Gray-Scott “zoo” discovered by Pearson (1993).

The Atlas

How to Read These

Each image shows the concentration field of chemical species V (the autocatalyst). Bright colors indicate high V concentration; dark areas are dominated by species U. The colormap runs from deep purple (v≈0) through orange to bright yellow (v≈0.5).

The same initial conditions are used for every simulation: a uniform field with a single square perturbation at the center. The spectacular diversity you see comes entirely from changing two numbers — F and k.

Pattern Taxonomy

PatternFkMechanism
Self-Replicating Spots0.03670.0649Spots grow, elongate, pinch, and split — like cell division
Maze / Labyrinth0.0290.057Stripes interconnect into branching networks
Coral / Chaos0.0200.054Filamentous branching growth — never stabilizes
Fingerprints0.05450.062Parallel stripe arrays that self-organize
Spot Lattice0.0220.051Quasi-hexagonal packing of stable spots
Holes / Inverse Spots0.0390.058Negaton-dominated: V everywhere except circular holes
Pulsating Spots0.0250.060Spots that breathe — oscillating in amplitude
Worms0.0460.0631Elongated stripes that grow from each end
Stripe Filaments0.0620.061Thin branching lines — the edge of pattern existence

More from Herman — see this project on the main site, alongside other work like Pokemon Emerald: Cathode, the Bobiverse novel, and the dashboard platform.