Home
Generated with Grok Build: Grok 4.5 · xAI Imagine (/imagine)
Biologically grounded Rust SNN primitives — LIF/Izhikevich SpikingNetwork, neuromodulators, STDP building blocks, and standalone neuron models.
| Version | 0.5.2 (release notes) |
| Tip (main) |
328ea9c · tag v0.5.2 = ef9fba8
|
| MSRV | 1.97.1 · Edition 2024 |
| License | MIT OR Apache-2.0 |
| Repo | https://github.com/Limen-Neural/neuromod |
| docs.rs | https://docs.rs/neuromod |
Topology-neutral, dynamically sized SpikingNetwork with strict step validation. The engine integrates LIF + Izhikevich banks only; Lapicque, GIF, FitzHugh–Nagumo, and Hodgkin–Huxley ship as standalone types. Four-channel neuromodulators, classical STDP and reward-modulated STDP. Pure computation library — no I/O, hardware, or training orchestration.
| Area | Capability |
|---|---|
| Network |
SpikingNetwork::new / with_dimensions, step → Result<Vec<usize>, StepError> (LIF + Izh banks) |
| Standalone neurons | Lapicque, GIF, FHN, HH (not alternate engine banks) |
| Modulators | DA, 5-HT, ACh, NE + GenericReward / SignalProfile
|
| Plasticity | Classical Hebbian STDP, R-STDP building blocks (eligibility types present; live engine path is direct DA-gated STDP) |
| Quality | Criterion benches (harness = false), multi-OS CI, Codecov, Docker Hub + GHCR |
| Page | Description |
|---|---|
| Release Notes | 0.5.2 + history |
| Overview | Cubic + Devin map |
| Getting Started | Install & quick start |
| Architecture | Engine and banks |
| Ecosystem Boundaries | LIM-9 owns / does not own |
| SpikingNetwork | step contract |
| Neuron Models | Model suite |
| Neuromodulators | DA/SE/ACh/NE API |
| Plasticity | STDP / R-STDP |
| Poisson Encoding | In-crate encoder note |
| Benchmarks | Criterion suite |
| CI and Quality | Workflows |
| Sentry | App-owned observability (no library feature) |
| Ecosystem | Sibling crates |
| Glossary | Terms |
use neuromod::{NeuroModulators, SpikingNetwork};
let mut network = SpikingNetwork::new(); // 16 LIF, 5 Izh, 16 channels
let stimuli = [0.5_f32; 16];
let modulators = NeuroModulators::default();
let spikes = network.step(&stimuli, &modulators).unwrap();[dependencies]
neuromod = "0.5.2"Dual-licensed under MIT or Apache-2.0 at your option.
Last updated: August 12, 2026
Updated by: Grok Build: Grok 4.5 (high)
Package tip reference: 328ea9c (main @ 0.5.2 + #97/#98; tag v0.5.2 = ef9fba8)
Cubic: page-intro
Devin DeepWiki: neuromod @ 328ea9c