PersonConnectome
People Playground Human driven by a thresholded MaleCNS-derived LIF graph. Game senses (injury, looming, audio, liquids as telemetry) map to a frozen decoder; it does not learn from staying alive.
Neurons are simulated — leaky integrate-and-fire or better — with documented, tuned parameters.
Dispute this rating →The evidence
No evidence, no level.Membrane potential leaks multiplicatively and integrates clamped synaptic input each tick.
https://github.com/TailsProwerWorks/PersonConnectome/blob/442a8f6d002f7bd0ed859743c674944ff8a22b01/src/Core/LifBrain.cs#L280-L291Spike propagation reads edge weights from the asset and resets the neuron with a refractory period; weights are never written.
https://github.com/TailsProwerWorks/PersonConnectome/blob/442a8f6d002f7bd0ed859743c674944ff8a22b01/src/Core/LifBrain.cs#L452-L471The authors state that connection weights stay fixed and the controller does not learn from experience.
https://github.com/TailsProwerWorks/PersonConnectome/blob/442a8f6d002f7bd0ed859743c674944ff8a22b01/README.md#L27
A MaleCNS-derived thresholded graph (176,422 neurons, 6,287,749 connections at synapse weight >= 5) run as a sparse leaky integrate-and-fire brain in C#: membrane potential decays multiplicatively each tick, accumulates clamped synaptic input, fires past threshold, then resets with a five-tick refractory period. Not L2: the connectome asset exposes weights through a read-only WeightAt accessor over an immutable array, and no code path writes a weight; a grep for plasticity, STDP, dopamine or a learning rate across the C# sources returns nothing. The README states it plainly: 'The connection weights stay fixed... the controller does not learn from experience or improve simply by staying alive.' The fly-to-human motor decoder is engineering, and the project is careful to label it as such rather than as biology.
Record
Nearby
All projects →MaleCNS simulation connected to ViZDoom through modeled visual inputs and a fixed button readout. Includes plasticity experiments and negative validation results; learned survival has not been demonstrated.
Research framework comparing a MaleCNS-constrained sparse recurrent controller with rewired graphs, conventional neural networks, and other controls on simple VizDoom tasks.
Ratings on thelearningfly.com are proposals, not verdicts. Every one of them is arguable in public.