Quantum + colony. Every new pump.fun pair, live, and 300 ants picking which to check.

New pairs launch faster than anyone can read them. The qolony decides which ones get a look, an ant reads the coin, its creator and its holders, and writes down what it finds. Live data, read only, and a report is a list of facts, not a rating.

nest
300ants
0new pairs seen
0checks carried out
0pairs on the field now

The field · new pump.fun pairs, live

waiting for the first new pairs…
every new pair, shown by its pictureblack frame = has a reportorange frame, bigger = still moving (sin²θ)ant heading out or readingant carrying a finding home
click any pair for its reportshowing the pair the colony is watching most

Report

no reports yet

Read budget 0 / 60 a minute

pump.fun allows about 60 reads a minute, so the colony has to choose where to spend them.

Colony state 300 ants

heading out0on a pair0heading home0in the nest0

Most watched P = sin²θ

nothing is moving yet

Every new pair · newest first

PairAgeMarket capSince seenCheckedsin²θ
Waiting for the first new pairs…
01 · reports

What the ants found

One row per pair that has had at least one look, the most watched first. Every cell is something an ant read from pump.fun's public data a moment ago. A dash means that read has not happened yet.

PairAgeMarket capSince seenLinksCreatorHoldersNotedsin²θ
The first reports will appear here within a few seconds.

Left the field

Pairs older than 20 minutes drop off the field. Their last report is kept here.

PairMarket capPeakLinksCreatorHoldersNoted
Nothing has left the field yet.
02 · method

How the colony chooses

The base is the Ant System (Dorigo, 1996), written from what real ants do at a forked bridge (Goss, Deneubourg and others, 1989). On top sits the quantum ant colony idea (Han and Kim, 2002; Wang, Niu and Shen, 2007): the trail on each coin is a qubit-style state, an angle θ. Choosing a coin is a measurement of it, and a coin seen still moving is turned up with a rotation gate.

state of a pair|ψ⟩ = cos θ |0⟩ + sin θ |1⟩ θ = 0 nobody is watching, θ = π/2 everybody is
choose a pairP(pair) ∝ η(age) · ( 1 + 30 · sin²θ )
freshnessη(age) = 1 + 6 · e^( − age / 90 s ) a brand-new pair smells strongest
rotation gateθ ← min( π/2, θ + 0.1 · signs of life ) market cap up 3%, new replies, new trades
decoherenceθ ← ( 1 − 0.006 ) · θ every step, on every pair
what an ant doesone read: the coin's page, then its creator's other coins, then its holder list

So a pair that keeps moving keeps getting looked at, and a pair that dies on arrival is quietly dropped. Quantum-inspired means the maths of a qubit, run on an ordinary computer. There is no quantum hardware here and no quantum speed-up is claimed.

03 · parameters

Every number it runs on

ParameterSymbolValueMeaning
04 · real and ours

Where the line is

The ants are simulated and the "quantum" part is borrowed maths. The data is not simulated.

Real

  • Every pair on the field is a real coin created on pump.fun in the last 20 minutes, read from its public feed.
  • Market caps, replies, links, a creator's other coins and holder counts are what pump.fun returned at the time shown.
  • "Dev", "sniper" and "bundler" are pump.fun's own labels on holders. We repeat them; we do not work them out.
  • The choosing rule is a published variant of ant colony optimisation.

Ours

  • It is quantum-inspired, not quantum: ordinary code, no quantum computer, no claimed speed-up.
  • What counts as a sign of life, and every parameter value.
  • The 300 ants on screen show where attention is going. Only about one a second carries out a real read, because that is what pump.fun allows. The speech tags in the header mark those real reads as they happen.
  • A report is a list of facts. It is not a score, a rating or advice, and a clean report does not make a coin safe.
  • Read only: nothing is bought, sold, signed or posted. Every new pair pump.fun lists is shown except ones it flags as adult or that have slurs in the name. Pictures are the coins' own, loaded from pump.fun; we do not pick or check them.
05 · log

Ant log

06 · sources

Sources

Goss, Aron, Deneubourg, Pasteels. Self-organized shortcuts in the Argentine ant. Naturwissenschaften, 1989.

Dorigo, Maniezzo, Colorni. Ant system: optimization by a colony of cooperating agents. IEEE Transactions on Systems, Man, and Cybernetics, 1996.

Han, Kim. Quantum-inspired evolutionary algorithm for a class of combinatorial optimization. IEEE Transactions on Evolutionary Computation, 2002.

Wang, Niu, Shen. A novel quantum ant colony optimization algorithm. Bio-Inspired Computational Intelligence and Applications, Springer, 2007.

Data: pump.fun's public feed, read about once a second and never written to.