BRAINLABThe graph
65 regional nodes, linked by connections reported in tract-tracing studies. The file holds node IDs and edges — no recordings from individual neurons.
02 / RESEARCHPUBLISHED WIRING · EXPERIMENTAL MODEL
A feline wiring map.
A place to start
computing.
BrainLab takes an openly published graph of connections between regions of the cat cortex and uses it as the fixed skeleton of a small computational experiment.
65CORTICAL REGIONS
1,139DIRECTED LINKS
08RESERVOIR UPDATES
01FIXED READOUT
From wiring
to computation.
Try the model ↗The reservoir
Five normalised features feed fixed input weights. Synthetic activity then spreads along the published links for eight updates.
hᵢ′ = tanh(Wᵢx + 0.65 · mean(hⱼ))
The readout
A logistic layer squeezes the final state into one score. Its weights come from a seeded generator and are published with the graph — nothing here has been trained on market history.
s = σ(b + wᵀh)
MODEL CARD / V0.1What is inside.
What is inside.
What is not.
The whole experiment is small enough to list on one card. Nothing is hidden behind a server, and nothing is claimed that the numbers cannot support.
Run the computation yourself ↗Regions (nodes)65
Directed links (edges)1,139
Input features5
Fixed parameters · input weights + readout + bias391
Trained on market dataNo
The score is uncalibrated and has not been evaluated against any market. Treat it as a demonstration of the method, never as a signal to trade on.
What the model
really means.
Is this a living feline brain? −
It uses a published graph of how regions connect. The activity running over it is artificial. It is not a biological simulation, not a model of individual neurons, and not a reconstruction of how a cat thinks.
Which parts can change? +
Only the inputs you set in the lab. The wiring, the reservoir rule, the input weights and the readout are all fixed. The lab runs entirely in your browser and never fetches prices.
What do the animations show? +
The lab graph draws the actual 65 nodes and 1,139 links on an abstract sphere, not an anatomical layout. Node colours are the computed activations. The cat portrait and the decorative particles are artwork.