Artificial Immune System: Clonal Selection Optimizer

Landscape

Population Controls

Population size 40

Clones per elite 6

Mutation scale β 1.0

Replacement rate (diversity) 15%

Live Stats

Generation

Best affinity

0.000

Mean affinity

0.000

Best (x, y)

—

How it works

This is the Clonal Selection Algorithm (CLONALG), a bio-inspired computing technique modelled on how the adaptive immune system affinity-matures antibodies against an antigen. Instead of a real antigen, each antibody is a point (x, y) scored against a fitness landscape f(x, y); "affinity" is defined so higher = better:

affinity(x, y) = 1 / (1 + f(x, y))

Each generation:

Evaluate every antibody's affinity against the landscape.
Select the top-n highest-affinity antibodies (the elite).
Clone each elite antibody proportionally — better antibodies get more clones.
Hypermutate the clones with a rate inversely proportional to affinity: strong antibodies mutate gently (fine local search), weak ones mutate wildly (exploration). Mutation step ∝ β · exp(−ρ · affinity), the CLONALG hypermutation law.
Re-select the best clone of each family back into the population.
Replace the lowest-affinity fraction of the population with fresh random antibodies — the "receptor editing" step that keeps diversity from collapsing.

Watch the sphere cloud contract onto the landscape's peaks over generations — the same affinity-maturation dynamic your immune system uses to home in on a pathogen, repurposed here as a general-purpose optimizer for functions with many local optima.

Gen 0

Low affinity Mid affinity High affinity (elite)

⚠ Couldn't load the 3D engineThree.js failed to load from the CDN. Check your connection and reload.