Physarum polycephalum is a single-celled, multi-nucleate slime mold that forages by extending a network of tube-like protoplasmic veins across its environment. This simulation models that behavior with an agent-based system: thousands of simple "particles" sense a chemical trail field ahead of them, steer toward the strongest concentration, and deposit their own trail as they move. No agent has a global plan — the branching, web-like network you see emerges purely from these local sensing-and-depositing rules repeated over time, then reinforced or pruned as trails decay. This mirrors how the real organism is famous for solving maze-like transport problems, notably approximating the efficiency of the Tokyo rail network in a well-known 2010 lab experiment, and it remains a popular model system in swarm intelligence and unconventional/biological computing research.

• Real Physarum can navigate mazes and reliably find the shortest path between food sources.
• Each simulated agent has a sensor angle and sensor distance used to sample trail concentration to its left, right, and ahead.
• Agents turn toward whichever sensor reads the highest concentration, then deposit new trail chemical at their position.
• The trail field continuously decays and diffuses, which lets weak paths fade while strong paths self-reinforce.
• Large-scale simulations commonly use tens of thousands of agents; this scene uses a few thousand for real-time 3D rendering.
• The sense-and-follow rule set is closely related to ant colony optimization and other stigmergic (trail-based) swarm algorithms.
• The real organism can grow several centimeters per day while foraging for food.
• Physarum has no brain or neurons, yet exhibits memory-like and decision-making behavior through its cytoplasmic network.