Swarm Lab · Experiment 02 · DNA 没写的身体
A body without a blueprint
How much shape can a few local rules make?
Nobody receives a drawing of the finished body. Particles notice nearby particles, keep a little personal space, and sometimes stick together. Then we remove fourteen of them.
Your challenge
See whether local attraction reconnects a damaged cluster.
- Play through step 90, when 14 central particles disappear.
- Scrub backward and forward around 90. Count survivors, not just connections.
- Try zero adhesion, a shorter signal range, or no damage. Change one thing at a time.
First time here? The buttons explained
Play starts or pauses. Step advances one frame; the time slider rewinds. Changing a control prepares a new run; press Play again. Reset replays the current settings. Seed chooses a random world, not difficulty; keep it while changing one rule, then try Next seed. Share saves your setup in a link. The downloads let you inspect the data.
What every parameter means
- Seed
- A repeatable random world, from 0 to 4294967295. The compared arms use the same seed.
- Adhesion
- Adhesion sets the strength of attraction between nearby particles in the second arm. 0 means no attraction; 1 is the strongest available. Repulsion and noise remain.
- Boundary
- The wall particles cannot cross. Shape can influence the pattern.
- Signal range
- How far a particle can attract another, as a fraction of the unit arena width. The connectivity measuring threshold stays 0.085.
- Remove 14 at step 90
- If checked, remove the fourteen nearest the center in each arm. If unchecked, keep all 72.
Make a prediction. Then press Play.
Model 1.1.0 · simulation onlyEvery dot follows the same local force rule. Lines join neighbors within the connection threshold. After step 90, fourteen central particles are removed in each arm. The dashed chart line is the first arm. Controls change both arms except the named treatment; the seed is shared.
Read the result: The score is the largest connected group divided by survivors. Reconnection is not regrowth: removed particles never return.
Loading local simulation…
Now break the story
Switch off damage. Then shrink the signal range. Does the cluster need long-range communication?
Move one control, keep the seed, and compare again. The batch below is a fixed baseline—not a live summary of your current settings.
Observed in this model · 32 matched seeds · all runs retained
A result, not just an animation.
Adhesion usually pulls more surviving particles into a connected group. It does not regrow the missing fourteen. Not every seed ends in a single cluster.
| Arm | Mean | Seed range |
|---|---|---|
| No adhesion | 50.7% | 22.4% – 86.2% |
| Local adhesion | 96.9% | 51.7% – 100.0% |
Ranges describe these runs, not confidence intervals for real-world systems.
Under the hood · the actual rules
72 particles start at matched positions. At distance below 0.043 they repel; between 0.043 and the signal range they attract in the adhesion arm. Positions update simultaneously with small deterministic noise and a shared boundary. Connectivity uses distance < 0.085. Damage removes the fourteen particles nearest the dish center at step 90. No particle is created afterward.
Read this model’s source · Read invariant tests · Reproduce the receipts
Where the analogy stops
These are abstract particles, not biological cells. Connectivity is not intelligence, healing or a new organism. Any post-damage reconnection is rearrangement of survivors. The rules and boundary—not DNA or an AI—supply the structure.
Sources, methods & reproduction
Inspired by Robin’s diary. Background: Tufts · Anthrobots research overview. This is an original teaching model, not a replication of that research.
Experiment notes · Run locally with Node · Version & source hashes