MCPcopy Create free account
hub / github.com/BindsNET/bindsnet / CSRMNodes

Class CSRMNodes

bindsnet/network/nodes.py:1319–1552  ·  view source on GitHub ↗

A layer of Cumulative Spike Response Model (Gerstner and van Hemmen 1992, Gerstner et al. 1996) nodes. It accounts for a model where refractoriness and adaptation were modeled by the combined effects of the spike after potentials of several previous spikes, rather than only the most rec

Source from the content-addressed store, hash-verified

1317
1318
1319class CSRMNodes(Nodes):
1320 """
1321 A layer of Cumulative Spike Response Model (Gerstner and van Hemmen 1992, Gerstner et al. 1996) nodes.
1322 It accounts for a model where refractoriness and adaptation were modeled by the combined effects
1323 of the spike after potentials of several previous spikes, rather than only the most recent spike.
1324 """
1325
1326 def __init__(
1327 self,
1328 n: Optional[int] = None,
1329 shape: Optional[Iterable[int]] = None,
1330 traces: bool = False,
1331 traces_additive: bool = False,
1332 tc_trace: Union[float, torch.Tensor] = 20.0,
1333 trace_scale: Union[float, torch.Tensor] = 1.0,
1334 sum_input: bool = False,
1335 rest: Union[float, torch.Tensor] = -65.0,
1336 thresh: Union[float, torch.Tensor] = -52.0,
1337 responseKernel: str = "ExponentialKernel",
1338 refractoryKernel: str = "EtaKernel",
1339 tau: Union[float, torch.Tensor] = 1,
1340 res_window_size: Union[float, torch.Tensor] = 20,
1341 ref_window_size: Union[float, torch.Tensor] = 10,
1342 reset_const: Union[float, torch.Tensor] = 50,
1343 tc_decay: Union[float, torch.Tensor] = 100.0,
1344 theta_plus: Union[float, torch.Tensor] = 0.05,
1345 tc_theta_decay: Union[float, torch.Tensor] = 1e7,
1346 lbound: float = None,
1347 **kwargs,
1348 ) -> None:
1349 # language=rst
1350 """
1351 Instantiates a layer of Cumulative Spike Response Model nodes.
1352
1353 :param n: The number of neurons in the layer.
1354 :param shape: The dimensionality of the layer.
1355 :param traces: Whether to record spike traces.
1356 :param traces_additive: Whether to record spike traces additively.
1357 :param tc_trace: Time constant of spike trace decay.
1358 :param trace_scale: Scaling factor for spike trace.
1359 :param sum_input: Whether to sum all inputs.
1360 :param rest: Resting membrane voltage.
1361 :param thresh: Spike threshold voltage.
1362 :param refrac: Refractory (non-firing) period of the neuron.
1363 :param tc_decay: Time constant of neuron voltage decay.
1364 :param theta_plus: Voltage increase of threshold after spiking.
1365 :param tc_theta_decay: Time constant of adaptive threshold decay.
1366 :param lbound: Lower bound of the voltage.
1367 """
1368 super().__init__(
1369 n=n,
1370 shape=shape,
1371 traces=traces,
1372 traces_additive=traces_additive,
1373 tc_trace=tc_trace,
1374 trace_scale=trace_scale,
1375 sum_input=sum_input,
1376 )

Callers 1

test_post_preMethod · 0.90

Calls

no outgoing calls

Tested by 1

test_post_preMethod · 0.72