Observe. Connect.
Make it your own.

One Python session loop connects neural activity to a small game environment or synthesized sound. Run a preset, inspect the inputs and controls, then replace an adapter.

Full-brain voxel world · recorded GPU experiment

All 139,255 FlyWire neurons run on an A16 GPU and control a ground body through an explicit DNa02 readout. Compare the actual trajectory with a silenced-input control. It moves, but curves away from food; no successful navigation or training is claimed.

Watch the voxel experiment →

Full FlyWire brain · CUDA benchmark

All 139,255 proofread neurons and 16.85 million source rows on a real NVIDIA GPU. CPU/CUDA numerical checks, throughput, checkpoint replay and continuous simulation; this is a full-graph software benchmark, not a trained game policy.

Read the measured results →

Reflex School · reward-trained readout

A fixed real 313-neuron circuit, three learned external action weights, and an actual A16 GPU run. On 80 balanced held-out two-choice trials, accuracy changed from 50% to 100%. This is readout learning, not biological plasticity.

Compare before and after →

Godot connectome game adapter

No CUDA required. Run an actual Godot 4 scene with the Python CPU SDK. Godot owns obstacles and movement; a replaceable encoder and neural readout supply steering. Includes pause, paired world/brain saves, restore and stop-on-disconnect behavior.

git clone --branch v0.5.0a1 https://github.com/freeman-1984-coder/flybrain-sdk.git
cd flybrain-sdk
python -m pip install -e .
python examples/godot/bridge.py
# In a second terminal, from the same directory:
godot --path examples/godot

Alternatively import examples/godot/project.godot in Godot and press F5. Wait for the bridge to print its ready address, then click Run / pause. The artificial toy works offline after installation. For the real 313-cell model, restart the bridge with --model male-cns-escape-v1 --download.

The scene advances 20 ms per acknowledged action, independently of rendering. Save completes any outstanding action before writing a checkpoint. Restore both the brain and the world together. The local HTTP bridge is a development example; it is not a hosted multiplayer service.

Godot setup and adapter guide Get the editable scene

Verified with toy and real circuits against Python feedback traces and whole-brain checkpoints. Real anatomical wiring uses assumed LIF dynamics and engineered steering; no trained avoidance or biological fidelity is claimed.

Neural dodge

A player moves left or right as obstacles descend. Neural outputs determine steering; the environment reports the movement actually applied. Both recordings use the same seed, real 313-cell model and six-second duration.

Watch the neural run Compare silenced outputs

This single example is a causal comparison, not evidence of learned avoidance or general task performance. Input currents and GF-to-steering mappings are engineered.

Circuit tones

Two synthetic pulse streams stimulate the same model. Neural firing rates control the volume of original 440 Hz and 660 Hz tones.

Inspect recorded input and output gains → · Download WAV

This is sonification, not musical understanding or a trained rhythm-game policy. The sound is synthesized from applied output gains.

Generate and edit your own

pip install -e .
flybrain init my-fly --template dodge
cd my-fly
python app.py --output runs/first

The generated project contains editable recipe.json, app.py, pinned requirements and attribution. Change input gains or output weights, then rerun to a new directory. The app exports a checkpoint, verified recording and HTML playback. Use --template tones at creation for WAV output. Add --model male-cns-escape-v1 --download to flybrain init for a real model recipe.

from flybrain.demos import make_demo

session = make_demo("dodge")
frame = session.step()
print(frame.observation)
print(frame.requested, frame.applied)
session.save("session.json")

Check your installation with flybrain doctor; discover templates with flybrain demos list. These commands also work outside a repository checkout after installing the package.

Replace the parts that matter

Change the feature-to-current projection, the rate-to-action readout, or the environment while keeping the same neural runtime and checkpoint format. The session tracks fractional timing, pending inputs and environment state.

Read the integration guide Edit the example

Build manifest and checksums · Source model and assumptions · Contribute a kit

The live browser game is available in the sandbox. The Godot adapter is available above; additional engines and streaming audio remain future work. The separate experimental CUDA branch has passed 12 A16 hardware cases; see Reflex School for the measured report and external readout learning example.