I have created a biologically based AI model

About this article

I've spent the last year building NIMCP — a biologically-inspired artificial brain in C that trains six different neural network types simultaneously (spiking, liquid, convolutional, Fourier, Hamiltonian, adaptive) with gradient flow between them through learnable bridges. Some things that might be interesting to this crowd: - The SNN developed 26 Hz firing rates with 67% sparsity — within mammalian cortical range — without any regularization targeting those values. It emerged from cross-netw...