LIVE NEURAL EXPERIMENT Robinhood Chain · Awaiting token launch · Payouts not enabled.
✳FLYLOTTO
SMALL BRAIN. BIG POSSIBILITIES.

THE DRAW ROOM.

⟐
Built for Robinhood ChainPowered by the Pons creator-fee model
FLYLOTTO CONTRACT · ROBINHOOD MAINNETAwaiting contract address
PRE-LAUNCH

THE FLY ARENA

CONNECTING
RUN EXPLORE
Canonical grayscale arena seen by the fly model
X — · Y —
⌖ TRACKING FLY_01
Fruit fly with red eyes and translucent wingsFLY_01 ↗
Neural preview · awaiting holders00:00
Fly cam 01
NEURONS FIRING—
L1 VISUAL INPUT— Hz
MOTOR RESPONSEAwaiting model
CONNECTED NEURONS—
Connecting to PythonWaiting for actual telemetry
↗
CREATOR FEES FOR PRIZES100%
◎
ARENA HOLDERS—
✳
RECENT COMPLETED RUNS0
⤴
SNAPSHOT BLOCKAwaiting CA
THE FLIGHT LOG

Real runs. Recorded results.

NEURAL RUN LOG
RUNOUTCOMESTEPSRECORD
Connecting to the neural run log…
THE ACTUAL CONNECTOME

The circuit
is connected.

Unmodified flycoinrh code turns visual input into neural spikes. Its motor output now moves this fly.

YOUR FLYLOTTO ENTRIES

Your balance. Your share.

Look up your wallet in the arena’s saved holder snapshot.

Awaiting holder data—

No signature, transfer, or token approval is requested.
BALANCE-WEIGHTED ENTRIES

Every eligible holder is included.

Entry weight is a wallet’s token balance divided by the eligible token balance in one confirmed snapshot. Treasury, burn, and verified liquidity-system addresses are excluded.

Up to 24 holder balls are shown. With more holders, balls represent groups covering the entire eligible snapshot. Select a ball to inspect its members.

Ball size reflects balance share. The neural model has spatial biases, so entry weight is not a verified winning probability. A selected group is not automatically assigned a wallet winner. Prize payouts remain disabled.
CONFIRMED HOLDER SNAPSHOT

Holder entries

Awaiting the token contract address.

HOLDERBALANCEENTRY WEIGHT
No token is connected yet.

A run keeps its starting snapshot fixed. Holder changes are applied between runs.
INSIDE FLYLOTTO

The brain is in the loop.

The arena is connected to the actual Python simulator from fruitflydev/flycoinrh.

  1. 01

    The model sees the arena

    A 1280 × 800 grayscale scene contains balls built from a confirmed holder snapshot. Before a token is connected, six preview targets are shown. Decorative labels and fly artwork are excluded from neural input.

  2. 02

    Neurons drive the fly

    Each call simulates 20 milliseconds of neural activity. DNa02 steering, DNa01 forward drive, MDN reverse, and DNp09 stopping determine movement and clicks. The upstream simulator resets neural membrane state between calls.

  3. 03

    A neural click can select a ball

    The first neural click inside a ball records an experimental holder or group candidate. A run ends after 100 steps if none is selected. There is no forced landing or fallback winner. Cursor movement wraps at the edges.

  4. 04

    Every run keeps its evidence

    The complete holder snapshot, block hash, arena mapping, model hashes, seeds, motor output, and outcome are frozen into the run record.

100% of creator fees are designated for holder prizes. Fees first accrue through Pons and are claimed to the creator-fee treasury. Holder snapshots connect after launch. Neural selections remain experimental; a verified payout draw and signed transfers are not enabled.
ACTUAL NEURAL RUN RECORD

Run —

Selection—
Recorded activity—
Outcome—

This record contains actual simulator output. It does not represent a holder winner, independently verified lottery, or payout transaction.
Download inputs and neural log ↓
MODEL PROVENANCE

Actual code. Actual spikes.

Waiting for the neural service…

Upstream commit—Connectome graph SHA-256—Simulation parameters

—

Open the pinned GitHub source ↗

The upstream FlyBrain, FlyEye, and FlyPilot modules are unmodified. The FLYLOTTO adapter supplies the scene, wrapped cursor boundary, and ball hit test. Learning is disabled.

Connectome data © HHMI Janelia FlyEM, Cambridge Connectomics Group, and Google Research · CC-BY 4.0. Code: MIT.

The Python simulator currently runs on the host computer. If that computer or its connection goes offline, the arena freezes and reports the loss of connection.