ferromorphic 0.15.0

Neuromorphic computing in pure Rust, the open corpus in one library, every module checked against a closed form and audited by mutation: spiking neurons from Hodgkin-Huxley to resonate-and-fire, synapses, STDP to three-factor and dendritic learning, surrogate gradients, BPTT, e-prop, exact spike-time gradients, ANN-to-SNN conversion, spiking convolutions and attention, reservoirs, the NEF, vector symbolic architectures, sparse coding, associative memories, QUBO by stochastic spiking, oscillators, ring attractors and neural fields, predictive coding, equilibrium propagation, a cerebellum, grid cells, learnable delays, stochastic resonance, spike-train metrics, graph algorithms by spike wavefront, cochlea, olfaction, touch, proprioception, sound localisation, event vision and fusion, hardware models, NIR, AEDAT 2-4/EVT3/N-MNIST/DVS-Gesture readers, NeuroBench metrics, and a joules ledger that prices synapse memory traffic. Zero dependencies, wasm-clean, deterministic by seed.