spiking-aeon — Package 26

Neuromorphic SNN Hardware Bridge (Intel Loihi 2) — GenesisAeon Entropy Atlas Package 26

Bridges the GenesisAeon CREP weight system to physical Spiking Neural Network hardware (Intel Loihi 2). CREP tensor components C, R, E, P are translated into LIF neuron parameters. Software simulation runs without hardware via the pure-Python Brian2-compatible backend.

Calibrated against NeuEdge (arXiv:2602.02439, 2026): 847 GOp/s/W · 2.3 ms latency · 89% core utilisation · 312x energy improvement over GPU.

SNN UTAC Model

CREP -> LIF Parameter Mapping

Install

pip install spiking-aeon
# or
uv add spiking-aeon

Usage

from spiking_aeon.system import SpikingAeon

sim = SpikingAeon(n_neurons=1000, seed=42)

# Software simulation (no hardware required)
result = sim.run_cycle(duration_ms=1000.0)
print(f"Mean firing rate: {result['mean_rate_hz']:.1f} Hz")

# CREP criticality tensor
crep = sim.get_crep_state()
print(f"Gamma = {crep['Gamma']:.4f}  (reference: {crep['Gamma_ref']})")

# UTAC state
utac = sim.get_utac_state()
print(f"H = {utac['H']:.4f}, below threshold: {utac['below_threshold']}")

# Zenodo record
record = sim.to_zenodo_record()

Stochastic Resonance

sr = sim.optimise_stochastic_resonance(n_steps=20)
print(f"Optimal noise D_res = {sr['D_res']:.3f}  (min CV = {sr['min_CV']:.3f})")

Deploy to Loihi 2

# Requires INRC access + pip install lava-nc
# https://www.intel.com/content/www/us/en/research/neuromorphic-computing.html
sim.deploy_to_loihi()

Package Structure

src/spiking_aeon/
├── __init__.py              # version 0.1.0, gamma=0.150, package_number=26
├── constants.py             # Loihi 2 specs, LIF params, UTAC constants
├── lif_neuron.py            # LIF neuron with CREP-modulated firing threshold
├── stdp_plasticity.py       # Spike-Timing Dependent Plasticity
├── stochastic_resonance.py  # Optimal noise sweep (Ferreira 2025)
├── crep_to_weights.py       # CREP {C,R,E,P} -> {tau_m, noise, weight, tau_ref}
├── crep_snn.py              # SNN CREP tensor, Gamma_ref ~= 0.150
├── brian2_backend.py        # Pure-Python LIF network simulation
├── loihi_adapter.py         # Lava SDK interface (requires INRC)
├── system.py                # Diamond interface (run_cycle, get_crep_state, ...)
└── benchmark.py             # NeuEdge 2026 benchmark suite

CREP Criticality Spectrum Position

Domain                       Pkg   Gamma    eta
---------------------------  ----  -------  ----
Qubit decoherence (T2)       P24   0.050    ~5%
Apoptosis (ATP threshold)    P25   0.090    20%
Amazon Rainforest            P19   0.116    12%
SNN firing (Loihi 2)         P26   0.150    32%   <- spiking-aeon
Seismic b=1.5 (GR law)       P23   0.200    40%
AMOC / Neural criticality    P20   0.251    50%
BTW Sandpile (SOC)           P22   0.296    58%

Role in the GenesisAeon Ecosystem

spiking-aeon is Package 26 of the GenesisAeon Entropy Atlas, in the neuromorphic hardware domain. It bridges the GenesisAeon CREP criticality framework to physical Spiking Neural Network hardware (Intel Loihi 2), positioning SNN firing-rate criticality (Γ ≈ 0.150) on the ecosystem's broader CREP Criticality Spectrum alongside other domains such as qubit decoherence, apoptosis, and self-organized criticality.

License

This repository is dual-licensed:

• Code (src/, tests/) — GNU GPLv3-or-later
• Documentation (README, docs/, RELEASE_GUIDE.md, CONTRIBUTING.md, and other prose) — Creative Commons Attribution 4.0 International (CC BY 4.0)

Citation

DOI will be updated automatically on each GitHub Release once Zenodo–GitHub integration is enabled for this repo.

@software{romer2026spiking_aeon,
  author    = {Romer, Johann},
  title     = {spiking-aeon: Neuromorphic SNN Hardware Bridge (Package 26)},
  year      = {2026},
  version   = {1.0.0},
  publisher = {Zenodo},
  doi       = {10.5281/zenodo.19645351},
}

@misc{neuredge2026,
  title  = {NeuEdge: Energy-Efficient Neuromorphic Edge Computing},
  year   = {2026},
  eprint = {2602.02439},
  note   = {847 GOp/s/W on Loihi 2},
}