spark.nn.components.synapses.linear

Classes

LinearSynapsesConfig	LinearSynapses model configuration class.
LinearSynapses	Linea synaptic model.

Module Contents

class spark.nn.components.synapses.linear.LinearSynapsesConfig(__skip_validation__=False, **kwargs)

Bases: spark.nn.components.synapses.base.SynanpsesConfig

LinearSynapses model configuration class.

Parameters:

__skip_validation__ (bool)

units: tuple[int, Ellipsis]

kernel: jax.Array | spark.nn.initializers.base.Initializer

class spark.nn.components.synapses.linear.LinearSynapses(config=None, **kwargs)

Bases: spark.nn.components.synapses.base.Synanpses

Linea synaptic model. Output currents are computed as the dot product of the kernel with the input spikes.

Init:

units: tuple[int, …] kernel: jax.Array | Initializer

Input:

spikes: SpikeArray

Output:

currents: CurrentArray

Reference:

Neuronal Dynamics: From Single Neurons to Networks and Models of Cognition. Gerstner W, Kistler WM, Naud R, Paninski L. Chapter 1.3 Integrate-And-Fire Models https://neuronaldynamics.epfl.ch/online/Ch1.S3.html

Parameters:

config (LinearSynapses | None)

config: LinearSynapsesConfig

build(input_specs)

Build method.

Parameters:

input_specs (dict[str, spark.core.specs.PortSpecs])

get_kernel()

Return type:

spark.core.payloads.FloatArray

get_flat_kernel()

Return type:

spark.core.payloads.FloatArray

set_kernel(new_kernel)

Parameters:

new_kernel (spark.core.payloads.FloatArray)

Return type:

None