Doomfly
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.
Synapses change under a modeled rule grounded in fly biology — for example dopamine-gated change on Kenyon-cell → MBON synapses in the mushroom body. Learning begins here.
Dispute this rating →The evidence
No evidence, no level.Damage schedules an aversive input into two PPL101 dopamine cells, and KC plus dopamine activity drive a plasticity rule on 4,184 existing KC→MBON11 connections while the rest of the wiring stays fixed.
https://github.com/nftechie/doomfly/blob/71ecf53d78eaffaf1a57ed7b0ccf5d458abc9f33/README.mdThe plastic edges are selected from real positive-weight KC→MBON11 edges in MaleCNS v1.0, and the code refuses to run if the expected MBON11 and PPL101 cells are not present in the release.
https://github.com/nftechie/doomfly/blob/71ecf53d78eaffaf1a57ed7b0ccf5d458abc9f33/doom_learning/circuit.py#L6-L22The project reports its own negative result: 'live experimental training, not demonstrated learned survival … Changing weights and longer individual rounds do not establish learning.'
https://github.com/nftechie/doomfly/blob/71ecf53d78eaffaf1a57ed7b0ccf5d458abc9f33/README.md
L2 is awarded for a synaptic change rule that runs inside the fly circuit, not for demonstrated learning — the project states plainly that its v6 candidate failed its visual, conditioning and survival gates. That honesty is the reason the evidence here is strong. Not L3: no published comparison against real fly experiments.
Record
Nearby
All projects →Research framework comparing a MaleCNS-constrained sparse recurrent controller with rewired graphs, conventional neural networks, and other controls on simple VizDoom tasks.
Hooks a MaleCNS model up to Super Mario 64, with a local dashboard and macOS setup. A playful experimental controller; requires your own game ROM.
Ratings on thelearningfly.com are proposals, not verdicts. Every one of them is arguable in public.