Model directory

Model directory#

The NEST-compatible neuron models grouped by family. Each family links to the API reference for the exact signature and to the upstream NEST documentation for the biophysics.

We link out for physiology

This page is a directory, not a model textbook. We deliberately do not duplicate NEST’s model semantics — for the equations, parameter meanings, and references of any model, follow the link to the upstream NEST documentation. What lives here is the brainpy.state surface: which models exist, how they are grouped, and how each is integrated (see Integration categories).

All models are constructed through the Simulator:

import brainunit as u
from brainpy import state as bp

sim = bp.Simulator(dt=0.1 * u.ms)
pop = sim.create(bp.iaf_psc_alpha, 100, I_e=0. * u.pA)   # a population of 100

Parameters use NEST’s names and units (C_m in pF, tau_m in ms, V_th in mV, I_e in pA, …). Pass them as keywords, or as a params=dict(...) mapping for a large parameter set.

Integrate-and-fire — current-based (psc)#

Linear current-based IAF neurons; integrated with exact analytic propagators (Integration categories category B — the most numerically faithful family).

iaf_psc_delta, iaf_psc_delta_ps, iaf_psc_alpha, iaf_psc_alpha_multisynapse, iaf_psc_alpha_ps, iaf_psc_exp, iaf_psc_exp_multisynapse, iaf_psc_exp_htum, iaf_psc_exp_ps, iaf_psc_exp_ps_lossless.

The _ps (“precise spiking”) variants resolve the spike time between grid points; the _multisynapse variants accept multiple receptor ports.

Integrate-and-fire — conductance-based (cond)#

Conductance-based IAF neurons; integrated with the adaptive RKF45 step (category A/C).

iaf_cond_alpha, iaf_cond_alpha_mc, iaf_cond_beta, iaf_cond_exp, iaf_cond_exp_sfa_rr.

Specialized IAF variants: iaf_bw_2001, iaf_bw_2001_exact, iaf_chs_2007, iaf_chxk_2008, iaf_tum_2000.

Adaptive exponential IF (AdEx)#

Brette–Gerstner adaptive exponential neurons; adaptive RKF45 (category A).

aeif_cond_alpha, aeif_cond_alpha_astro, aeif_cond_alpha_multisynapse, aeif_cond_beta_multisynapse, aeif_cond_exp, aeif_psc_alpha, aeif_psc_delta, aeif_psc_delta_clopath, aeif_psc_exp.

aeif_psc_delta_clopath exposes the voltage filters the clopath_synapse reads — see STDP parity: where state lives and how spikes pair.

Generalized IF (GIF) and Generalized LIF (GLIF)#

GIF (Mensi/Pozzorini) and Allen-Institute GLIF neurons; adaptive RKF45 (category A).

GIF: gif_cond_exp, gif_cond_exp_multisynapse, gif_pop_psc_exp, gif_psc_exp, gif_psc_exp_multisynapse.

GLIF: glif_cond, glif_psc, glif_psc_double_alpha.

Multi-timescale adaptive threshold (MAT)#

mat2_psc_exp, amat2_psc_exp.

Hodgkin–Huxley family#

Biophysical conductance-based neurons; adaptive RKF45 (category C).

hh_psc_alpha, hh_psc_alpha_clopath, hh_psc_alpha_gap, hh_cond_exp_traub, hh_cond_beta_gap_traub, ht_neuron.

The _gap variants participate in gap-junction networks (see Connectivity).

Izhikevich and point-process neurons#

Izhikevich: izhikevich (adaptive RKF45, category A).

Point process: pp_psc_delta, pp_cond_exp_mc_urbanczik (the two-compartment neuron the Urbanczik dendritic-prediction rule reads).

Rate neurons#

Rate-based units; vectorized update (category D).

lin_rate_ipn, lin_rate_opn, tanh_rate_ipn, tanh_rate_opn, sigmoid_rate_ipn, sigmoid_rate_gg_1998_ipn, gauss_rate_ipn, threshold_lin_rate_ipn, threshold_lin_rate_opn, rate_neuron_ipn, rate_neuron_opn, rate_transformer_node, siegert_neuron.

Binary and other neurons#

Binary: mcculloch_pitts_neuron, ginzburg_neuron, erfc_neuron.

Relay / utility: parrot_neuron (relays its input spikes 1:1 — used to drive plastic edges or to tap a generator), ignore_and_fire (a fixed-rate source).

See also#