- NEST Users - NEST Initiative Mail

nestml synapses in pyNN
by Chris Huyck 14 Feb '25

14 Feb '25

I've taken your noisy_synapse model and got that to run in pyNest. When I try it in PyNN it fails (can't find weight if you set w, weight not an argument if you use weight). I understand that you don't support pyNN. Do you support SpiNNaker (as it's a target)? If so, how do you access the model on SpiNNaker? (I only access SpiNNaker through PyNN). -Chris

2 8

nestml doesn't work on synapse models
by Chris Huyck 12 Feb '25

12 Feb '25

I've got a new conda environment. I compile nest (3.7) from a tarball and it runs. I pip install pyNN I pip install nestml (which now gives me nestml8.0.0. Is there some way to tell the version without just uninstalling it?) The neuron models work. The synapse models fail to nestml I tried my models, but went to the noisy_synapse and static_synapse models (from git) Both fail in nest_code_generator.py line 267 (variables_special_cases = ...) Any thoughts about how to get synapses to work in nestml? -Chris

2 5

wrong results eprop examples
by Xavier Otazu Porter 11 Feb '25

11 Feb '25

Hello, I have downloaded several of the eprop python examples, and none of them generate the correct results. For example, the handwritten example, generates a rather random result (the red points data) that do not fit with the expected results. The sinus generation, also generates wrong results. If you want, I can send a small image (not did it now to avoid image spamming). I am using NEST 3.8 with python 3.9 in a linux Rocks enviornment, and python 3.11 in a FreeBSD environment. Thanks a lot in advance. Xavier Computer Vision Center<http://www.cvc.uab.cat> CONFIDENTIALITY WARNING<http://www.cvc.uab.es/?page_id=7475>

3 2

`tsodyks_synapse` and `tsodyks2_synapse` models have different behavior for short term depression
by Pavel Haman 10 Feb '25

10 Feb '25

Hello, I have followed the example script comparing `tsodyks_synapse` and `tsodyks2_synapse` models, which can be found here: https://nest-simulator.readthedocs.io/en/stable/auto_examples/evaluate_tsod… Using the parameter set for depression instead of facilitation I found that `tsodyks2_synapse` model leads to weird peaks of synaptic currents (at least I would not expect that for short term depression) - during the second stimulation the weight rescaling seems to be shifted by one spike compared to tsodyks_synapse (as if parameter "u" would be updated after rescaling the weight instead of before). Does anybody know why is there this difference between the two synapse models? Thank you. All the best, PH

3 12

Building a nest simulation for mnist dataset
by Aditya Srivastava 10 Feb '25

10 Feb '25

I have built a nest simulation for a binary dataset where we have two classes normal and anomalous . So i tried to built a model iun such a way so that the model learns to produce no spike in case of a normal data but when it sees a anomalous data it is producing the spikes. So this model was working fine. Now I am trying to move forward with it and try it on image dataset where I am using simple MNIST data and I have just taken image for one digit and trying to build a same kind of architecture where it should produce almost no spike when it sees a digit one and oin case of other digits it should produce spikes but I am facing issues in it . The model is not able to produce enough spikes and when I am creating more exc connections then spikes are not reducing with each epochs so can you please suggest me a better approach or some suggestion for my current approach. # loading dataset from sklearn.datasets import fetch_openml import numpy as np def preprocess_mnist_digit(digit=1): """ Load and preprocess the MNIST dataset to only include images of a specific digit. Normalize pixel values by dividing by 4 and flatten images. Args: digit (int): The digit to filter (e.g., 1 for digit '1'). num_samples (int): Number of samples to retain. Returns: flattened_images (array): Flattened MNIST images of the specified digit. labels (array): Corresponding labels. """ # Load MNIST data mnist = fetch_openml('mnist_784', version=1) data, labels = mnist['data'], mnist['target'].astype(int) # Convert labels to integers # Filter images that correspond to the specified digit digit_indices = np.where(labels == digit)[0] # Get indices where label is equal to the specified digit # Extract the corresponding images and labels filtered_images = data.iloc[digit_indices].to_numpy(dtype=float) / 4.0 # Normalize filtered_labels = labels[digit_indices] return filtered_images, filtered_labels # Load only images of the digit '1' images, labels = preprocess_mnist_digit(digit=1) # building simulation nest.ResetKernel() # Simulation parameters dt = 0.01 # Simulation time step in ms nest.SetKernelStatus({"resolution": dt}) N = 784 # Number of neurons images=images[:10] gmax = 3 neuron_params = { "tau_m": 140.0, # Membrane time constant in ms "E_L": -70.6, # Resting potential (in mV) "V_th": -40.4, # Threshold potential (in mV) "V_reset": -70.6, # Reset potential (in mV) "V_m":0.0, } nest.CopyModel("stdp_synapse", "stdp_synapse_exc", {"weight":gmax,"alpha":1,"Wmax":gmax, "tau_plus": 20.0,"lambda": -0.0005, "mu_minus": 0, "mu_plus": 0}) nest.CopyModel("stdp_synapse", "stdp_synapse_inh", {"weight":-gmax/(N-1),"alpha":1,"Wmax":-gmax, "tau_plus": 20.0,"lambda": 0.005, "mu_minus": 0, "mu_plus": 0}) G_spike_gen = nest.Create("poisson_generator",N) G_input = nest.Create("parrot_neuron", N) input_spike_recorder = nest.Create("spike_recorder",N) nest.Connect(G_input, input_spike_recorder,"one_to_one") G_hidden = nest.Create("iaf_psc_delta", N,params=neuron_params) G_output = nest.Create("iaf_psc_delta", 1,params=neuron_params) hidden_spike_recorder = nest.Create("spike_recorder",N) nest.Connect(G_hidden, hidden_spike_recorder,"one_to_one") S_in = nest.Connect(G_spike_gen, G_input, "one_to_one") # Connect input to hidden layer with excitatory STDP synapses S_exc = nest.Connect(G_input, G_hidden,"one_to_one",syn_spec = "stdp_synapse_exc") # Retrieve the neuron IDs for G_input and G_hidden input_ids = G_input.tolist() hidden_ids = G_hidden.tolist() # Get the existing excitatory connections and extract their source-target ID pairs exc_connections = nest.GetConnections(G_input, G_hidden) exc_connection_pairs = set(zip(exc_connections.get("source"), exc_connections.get("target"))) for i in input_ids: for h in hidden_ids: if (i, h) not in exc_connection_pairs: nc1 = nest.NodeCollection([i]) nc2 = nest.NodeCollection([h]) nest.Connect(nc1, nc2, syn_spec= "stdp_synapse_inh") S_out = nest.Connect(G_hidden, G_output,"all_to_all",syn_spec = {"weight": 40.4/N}) # training function def train_net(images, spike_tol=0): output_spike_recorder = nest.Create("spike_recorder") output_membrane_recorder = nest.Create("voltmeter", 1) nest.Connect(G_output, output_spike_recorder) # Record output layer spikes nest.Connect(output_membrane_recorder, G_output, "one_to_one") while True: for image in images: for i in range(784): G_spike_gen[i].rate=image[i] nest.Simulate(500) for i in range(784): spikes = nest.GetStatus(hidden_spike_recorder[i], "events")[0] # if length of spikes is not zero then print the number of spikes and the times of spikes also print the index of the input_spike_recorder if len(spikes["times"]) != 0: print("Number of spikes in hidden_spike_recorder",i,":", len(spikes["times"])) print("Times of spikes in hidden_spike_recorder",i,":", spikes["times"]) print("\n") output_spikes_num = len(output_spike_recorder.get()["events"]["times"]) print("# of output spikes: " + str(output_spikes_num)) # del recorder_for_epoch if output_spikes_num <= spike_tol: print("Training finished") break return output_membrane_recorder, output_spike_recorder v, spk = train_net(images, spike_tol = 2) ___________________________________________

2 1

Open NEST Developer Video Conference Monday February 10, at 11:30 CET (UTC+0100)
by Dennis Terhorst 10 Feb '25

10 Feb '25

Dear NEST Users & Developers! I would like to invite you to our next fortnightly Open NEST Developer Video Conference today *Monday February 10, at 11:30 CET (UTC+0100).* As usual, in the Project team round, a contact person of each team will give a short statement summarizing ongoing and planned work in the team and highlight cross-cutting points that need discussion among the teams. The remainder of the meeting we would go into a more in-depth discussion of topics that came up on the mailing list or that are suggested by the teams. Feel free to join the meeting also if it’s just to bring your own quick questions for direct discussion in the in-depth section. Agenda * Welcome * Review of NEST User Mailing List * Project team round * In-depth discussion The agenda for this meeting is also available online, see https://github.com/nest/nest-simulator/wiki/2025-02-10-Open-NEST-Developer-… <https://github.com/nest/nest-simulator/wiki/2025-02-10-Open-NEST-Developer-…> Looking forward to seeing you soon! Cheers, Dennis Terhorst Log-in information We use a virtual conference room provided by DFN <https://www.dfn.de/en/> (Deutsches Forschungsnetz). You can use the web client to connect. We however encourage everyone to use a headset for better audio quality or even a proper video conferencing system (see below) or software when available. Web client * Visit https://conf.dfn.de/webapp/conference/97938800 <https://conf.dfn.de/webapp/conference/97938800> * Enter your name and allow your browser to use camera and microphone * The conference does not need a PIN to join, just click join and you’re in. In case you see a dfnconf logo and the phrase “Auf den Meetingveranstalter warten”, just be patient, the meeting host needs to join first (a voice will tell you). VC system/software How to log in with a video conferencing system, depends on you VC system or software. * Using the H.323 protocol (eg Polycom): |vc.dfn.net##97938800| or |194.95.240.2##97938800| * Using the SIP protocol:97938800@vc.dfn.de <mailto:97938800@vc.dfn.de> * By telephone: |+49-30-200-97938800| For those who do not have a video conference system or suitable software, Polycom provides a pretty good free app for iOS and Android, so you can join from your tablet (Polycom RealPresence Mobile, available from AppStore/PlayStore). Note that firewalls may interfere with videoconferencing in various and sometimes confusing ways. For more technical information on logging in from various VC systems, please see http://vcc.zih.tu-dresden.de/index.php?linkid=1.1.3.4 <http://vcc.zih.tu-dresden.de/index.php?linkid=1.1.3.4> -- Dipl.-Phys. Dennis Terhorst Coordinator Software Development Institute for Advanced Simulation (IAS-6) Computational and Systems Neuroscience & JARA-Institute Brain Structure-Function Relationships (INM-10) Institute of Neuroscience and Medicine Jülich Research Center, Member of the Helmholz Association 52425 Jülich, Germany http://www.csn.fz-juelich.de/ Phone +49 2461 61-85062 ---------------------------------------------------------------------- Forschungszentrum Jülich GmbH 52425 Jülich Sitz der Gesellschaft: Jülich Eingetragen im Handelsregister des Amtsgerichts Düren Nr. HR B 3498 Vorsitzender des Aufsichtsrats: MinDir Stefan Müller Geschäftsführung: Prof. Dr. Astrid Lambrecht (Vorsitzende), Karsten Beneke (stellv. Vorsitzender), Prof. Dr. Ir. Pieter Jansens ----------------------------------------------------------------------

1 0

Invitation to EBRAINS events in Heidelberg from 12-13 March
by Dennis Terhorst 10 Feb '25

10 Feb '25

Hi everyone, I would like to forward the invitation to the EBRAINS events in Heidelberg next month. Many experts of tools will be there including quite a number of NEST users and developers. You have the opportunity to join tutorials, learn about more interesting tools and directly exchange experience with others. Have a look! --- Dear NEST Community, We are delighted to invite you to join us in Heidelberg, Germanyat theEINC <https://einc.eu/>, from 12-13 March 2025, for a series of EBRAINS events. On*Wednesday,**12 March*, we will host an*EBRAINS Tutorials and Users Day*. It is aimed at both newcomers and experienced users of EBRAINS tools. Newcomers will have the chance to explore the wide range of Open Science Tools and Services available on the research infrastructure, while seasoned EBRAINS users can dive deeper with advanced tutorials and discuss with developers. You can find out more about the Tutorials and Users Day here:https://www.ebrains.eu/news-and-events/accelerate-your-neuroscience-research-with-ebrains-tutorials-and-users-day-in-heidelberg <https://www.ebrains.eu/news-and-events/accelerate-your-neuroscience-researc…> On*Thursday, 13 March*, we will host an*EBRAINS Developers Day*. It will feature introductions for developers on how to integrate tools and services into the EBRAINS software distribution or to integrate services via EBRAINS (workflows). You can find out more about the Developers Day here:https://www.ebrains.eu/news-and-events/developers-day-2025 <https://www.ebrains.eu/news-and-events/developers-day-2025> If you have any questions, please don't hesitate to contactcommunications(a)ebrains.eu <mailto:communications@ebrains.eu>. We hope to see you in Heidelberg! Kind regards, EBRAINSAISBL --- -- Dipl.-Phys. Dennis Terhorst Coordinator Software Development Institute for Advanced Simulation (IAS-6) Computational and Systems Neuroscience & JARA-Institute Brain Structure-Function Relationships (INM-10) Institute of Neuroscience and Medicine Jülich Research Center, Member of the Helmholz Association 52425 Jülich, Germany http://www.csn.fz-juelich.de/ Phone +49 2461 61-85062 ---------------------------------------------------------------------- Forschungszentrum Jülich GmbH 52425 Jülich Sitz der Gesellschaft: Jülich Eingetragen im Handelsregister des Amtsgerichts Düren Nr. HR B 3498 Vorsitzender des Aufsichtsrats: MinDir Stefan Müller Geschäftsführung: Prof. Dr. Astrid Lambrecht (Vorsitzende), Karsten Beneke (stellv. Vorsitzender), Prof. Dr. Ir. Pieter Jansens ----------------------------------------------------------------------

1 0

Test
by Jochen Martin Eppler 10 Feb '25

10 Feb '25

1 1

Building a nest simulation for mnist dataset
by Aditya Srivastava 28 Jan '25

28 Jan '25

I have built a nest simulation for a binary dataset where we have two classes normal and anomalous . So i tried to built a model iun such a way so that the model learns to produce no spike in case of a normal data but when it sees a anomalous data it is producing the spikes. So this model was working fine. Now I am trying to move forward with it and try it on image dataset where I am using simple MNIST data and I have just taken image for one digit and trying to build a same kind of architecture where it should produce almost no spike when it sees a digit one and oin case of other digits it should produce spikes but I am facing issues in it . The model is not able to produce enough spikes and when I am creating more exc connections then spikes are not reducing with each epochs so can you please suggest me a better approach or some suggestion for my current approach. # loading dataset from sklearn.datasets import fetch_openml import numpy as np def preprocess_mnist_digit(digit=1): """ Load and preprocess the MNIST dataset to only include images of a specific digit. Normalize pixel values by dividing by 4 and flatten images. Args: digit (int): The digit to filter (e.g., 1 for digit '1'). num_samples (int): Number of samples to retain. Returns: flattened_images (array): Flattened MNIST images of the specified digit. labels (array): Corresponding labels. """ # Load MNIST data mnist = fetch_openml('mnist_784', version=1) data, labels = mnist['data'], mnist['target'].astype(int) # Convert labels to integers # Filter images that correspond to the specified digit digit_indices = np.where(labels == digit)[0] # Get indices where label is equal to the specified digit # Extract the corresponding images and labels filtered_images = data.iloc[digit_indices].to_numpy(dtype=float) / 4.0 # Normalize filtered_labels = labels[digit_indices] return filtered_images, filtered_labels # Load only images of the digit '1' images, labels = preprocess_mnist_digit(digit=1) # building simulation nest.ResetKernel() # Simulation parameters dt = 0.01 # Simulation time step in ms nest.SetKernelStatus({"resolution": dt}) N = 784 # Number of neurons images=images[:10] gmax = 3 neuron_params = { "tau_m": 140.0, # Membrane time constant in ms "E_L": -70.6, # Resting potential (in mV) "V_th": -40.4, # Threshold potential (in mV) "V_reset": -70.6, # Reset potential (in mV) "V_m":0.0, } nest.CopyModel("stdp_synapse", "stdp_synapse_exc", {"weight":gmax,"alpha":1,"Wmax":gmax, "tau_plus": 20.0,"lambda": -0.0005, "mu_minus": 0, "mu_plus": 0}) nest.CopyModel("stdp_synapse", "stdp_synapse_inh", {"weight":-gmax/(N-1),"alpha":1,"Wmax":-gmax, "tau_plus": 20.0,"lambda": 0.005, "mu_minus": 0, "mu_plus": 0}) G_spike_gen = nest.Create("poisson_generator",N) G_input = nest.Create("parrot_neuron", N) input_spike_recorder = nest.Create("spike_recorder",N) nest.Connect(G_input, input_spike_recorder,"one_to_one") G_hidden = nest.Create("iaf_psc_delta", N,params=neuron_params) G_output = nest.Create("iaf_psc_delta", 1,params=neuron_params) hidden_spike_recorder = nest.Create("spike_recorder",N) nest.Connect(G_hidden, hidden_spike_recorder,"one_to_one") S_in = nest.Connect(G_spike_gen, G_input, "one_to_one") # Connect input to hidden layer with excitatory STDP synapses S_exc = nest.Connect(G_input, G_hidden,"one_to_one",syn_spec = "stdp_synapse_exc") # Retrieve the neuron IDs for G_input and G_hidden input_ids = G_input.tolist() hidden_ids = G_hidden.tolist() # Get the existing excitatory connections and extract their source-target ID pairs exc_connections = nest.GetConnections(G_input, G_hidden) exc_connection_pairs = set(zip(exc_connections.get("source"), exc_connections.get("target"))) for i in input_ids: for h in hidden_ids: if (i, h) not in exc_connection_pairs: nc1 = nest.NodeCollection([i]) nc2 = nest.NodeCollection([h]) nest.Connect(nc1, nc2, syn_spec= "stdp_synapse_inh") S_out = nest.Connect(G_hidden, G_output,"all_to_all",syn_spec = {"weight": 40.4/N}) # training function def train_net(images, spike_tol=0): output_spike_recorder = nest.Create("spike_recorder") output_membrane_recorder = nest.Create("voltmeter", 1) nest.Connect(G_output, output_spike_recorder) # Record output layer spikes nest.Connect(output_membrane_recorder, G_output, "one_to_one") while True: for image in images: for i in range(784): G_spike_gen[i].rate=image[i] nest.Simulate(500) for i in range(784): spikes = nest.GetStatus(hidden_spike_recorder[i], "events")[0] # if length of spikes is not zero then print the number of spikes and the times of spikes also print the index of the input_spike_recorder if len(spikes["times"]) != 0: print("Number of spikes in hidden_spike_recorder",i,":", len(spikes["times"])) print("Times of spikes in hidden_spike_recorder",i,":", spikes["times"]) print("\n") output_spikes_num = len(output_spike_recorder.get()["events"]["times"]) print("# of output spikes: " + str(output_spikes_num)) # del recorder_for_epoch if output_spikes_num <= spike_tol: print("Training finished") break return output_membrane_recorder, output_spike_recorder v, spk = train_net(images, spike_tol = 2)

1 0

Open NEST Developer Video Conference Monday January 27, at 11:30 CET (UTC+0100)
by Dennis Terhorst 27 Jan '25

27 Jan '25

Dear NEST Users & Developers! I would like to invite you to our next fortnightly Open NEST Developer Video Conference today *Monday January 27, at 11:30 CET (UTC+0100).* As usual, in the Project team round, a contact person of each team will give a short statement summarizing ongoing and planned work in the team and highlight cross-cutting points that need discussion among the teams. The remainder of the meeting we would go into a more in-depth discussion of topics that came up on the mailing list or that are suggested by the teams. Feel free to join the meeting also if it’s just to bring your own quick questions for direct discussion in the in-depth section. Agenda * Welcome o Welcome NEAT * Review of NEST User Mailing List * Project team round * In-depth discussion The agenda for this meeting is also available online, see https://github.com/nest/nest-simulator/wiki/2025-01-27-Open-NEST-Developer-… <https://github.com/nest/nest-simulator/wiki/2025-01-27-Open-NEST-Developer-…> Looking forward to seeing you soon! Cheers, Dennis Terhorst Log-in information We use a virtual conference room provided by DFN <https://www.dfn.de/en/> (Deutsches Forschungsnetz). You can use the web client to connect. We however encourage everyone to use a headset for better audio quality or even a proper video conferencing system (see below) or software when available. Web client * Visit https://conf.dfn.de/webapp/conference/97938800 <https://conf.dfn.de/webapp/conference/97938800> * Enter your name and allow your browser to use camera and microphone * The conference does not need a PIN to join, just click join and you’re in. In case you see a dfnconf logo and the phrase “Auf den Meetingveranstalter warten”, just be patient, the meeting host needs to join first (a voice will tell you). VC system/software How to log in with a video conferencing system, depends on you VC system or software. * Using the H.323 protocol (eg Polycom): |vc.dfn.net##97938800| or |194.95.240.2##97938800| * Using the SIP protocol:97938800@vc.dfn.de <mailto:97938800@vc.dfn.de> * By telephone: |+49-30-200-97938800| For those who do not have a video conference system or suitable software, Polycom provides a pretty good free app for iOS and Android, so you can join from your tablet (Polycom RealPresence Mobile, available from AppStore/PlayStore). Note that firewalls may interfere with videoconferencing in various and sometimes confusing ways. For more technical information on logging in from various VC systems, please see http://vcc.zih.tu-dresden.de/index.php?linkid=1.1.3.4 <http://vcc.zih.tu-dresden.de/index.php?linkid=1.1.3.4> -- Dipl.-Phys. Dennis Terhorst Coordinator Software Development Institute for Advanced Simulation (IAS-6) Computational and Systems Neuroscience & JARA-Institute Brain Structure-Function Relationships (INM-10) Institute of Neuroscience and Medicine Jülich Research Center, Member of the Helmholz Association 52425 Jülich, Germany http://www.csn.fz-juelich.de/ Phone +49 2461 61-85062 ---------------------------------------------------------------------- Forschungszentrum Jülich GmbH 52425 Jülich Sitz der Gesellschaft: Jülich Eingetragen im Handelsregister des Amtsgerichts Düren Nr. HR B 3498 Vorsitzender des Aufsichtsrats: MinDir Stefan Müller Geschäftsführung: Prof. Dr. Astrid Lambrecht (Vorsitzende), Karsten Beneke (stellv. Vorsitzender), Prof. Dr. Ir. Pieter Jansens ----------------------------------------------------------------------

1 0

2025

2024

2023

2022

2021

2020

2019

NEST Users