Some days are better than others! Today, two papers by Arman Ferdowsi and me went online within hours. One introduces a new learning rule for neuromorphic computing, the other shows how to distribute it across edge devices. The common thread is sparsity, both inside a chip and on the network.
September 24, 2026 at 10:20 PM
Some days are better than others! Today, two papers by Arman Ferdowsi and me went online within hours. One introduces a new learning rule for neuromorphic computing, the other shows how to distribute it across edge devices. The common thread is sparsity, both inside a chip and on the network.
Our work on the #interpretability of #Capsule Networks have been accepted at #Neurocomputing. #sqIRL #UAntwerp
#CapsNets #XAI #AI #ML #DNN #ANN
#CapsNets #XAI #AI #ML #DNN #ANN
Congratulations to Saja Tawalbeh on the acceptance of her work on the #interpretability of #Capsule Networks at the #Neurocomputing Journal.
Pre-proof: www.sciencedirect.com/science/arti...
@uantwerpen.bsky.social #XAI #ML #AI #sqIRL
Pre-proof: www.sciencedirect.com/science/arti...
@uantwerpen.bsky.social #XAI #ML #AI #sqIRL
Towards the characterization of representations learned via capsule-based network architectures
Capsule Neural Networks (CapsNets) have been re-introduced as a more compact and interpretable alternative to standard deep neural networks. While rec…
www.sciencedirect.com
December 4, 2024 at 1:53 PM
Our work with Karl Friston on Self-Orthogonalizing Attractor Neural Networks is now out in Neurocomputing!
What does this theoretical model mean for our understanding of the brain? I’ve mapped out the key neuroscience implications below.
Read the thread for a neuroscience walk-through ↓
What does this theoretical model mean for our understanding of the brain? I’ve mapped out the key neuroscience implications below.
Read the thread for a neuroscience walk-through ↓
April 17, 2026 at 10:47 AM
Our work with Karl Friston on Self-Orthogonalizing Attractor Neural Networks is now out in Neurocomputing!
What does this theoretical model mean for our understanding of the brain? I’ve mapped out the key neuroscience implications below.
Read the thread for a neuroscience walk-through ↓
What does this theoretical model mean for our understanding of the brain? I’ve mapped out the key neuroscience implications below.
Read the thread for a neuroscience walk-through ↓
We map brain-wide functions of awake mice at the single-voxel scale !
Introducing PyfUS our brand new, ground-setting, open-source software for analysing brain-wide functional ultrasound imaging datasets.
Now in #Neurocomputing
@sciencedirect.bsky.social
doi.org/10.1016/j.ne...
Introducing PyfUS our brand new, ground-setting, open-source software for analysing brain-wide functional ultrasound imaging datasets.
Now in #Neurocomputing
@sciencedirect.bsky.social
doi.org/10.1016/j.ne...
July 16, 2025 at 10:44 AM
We map brain-wide functions of awake mice at the single-voxel scale !
Introducing PyfUS our brand new, ground-setting, open-source software for analysing brain-wide functional ultrasound imaging datasets.
Now in #Neurocomputing
@sciencedirect.bsky.social
doi.org/10.1016/j.ne...
Introducing PyfUS our brand new, ground-setting, open-source software for analysing brain-wide functional ultrasound imaging datasets.
Now in #Neurocomputing
@sciencedirect.bsky.social
doi.org/10.1016/j.ne...
I read this in graduate school when I was sharper and probably got < 20%.
ieeexplore.ieee.org/document/58324
ieeexplore.ieee.org/document/58324
Mathematical foundations of neurocomputing
An attempt is made to establish a mathematical theory that shows the intrinsic mechanisms, capabilities, and limitations of information processing by various architectures of neural networks. A method...
ieeexplore.ieee.org
November 23, 2025 at 6:08 PM
I read this in graduate school when I was sharper and probably got < 20%.
ieeexplore.ieee.org/document/58324
ieeexplore.ieee.org/document/58324
"From silicon to spikes: ..." (Neurocomputing) computes exact gradients for weights, delays, and thresholds, directly at spike times. This reduces the estimated on-chip memory traffic by 24x compared to a standard surrogate-gradient baseline. doi.org/10.1016/j.neucom.2026.135225
September 24, 2026 at 10:20 PM
"From silicon to spikes: ..." (Neurocomputing) computes exact gradients for weights, delays, and thresholds, directly at spike times. This reduces the estimated on-chip memory traffic by 24x compared to a standard surrogate-gradient baseline. doi.org/10.1016/j.neucom.2026.135225
New work led by Filippo out on Neurocomputing! W/ @ksks.bsky.social , @kanair.bsky.social and me.
Does an agent need to be "aware" of its executed actions to learn to achieve a certain goal?
In reinforcement learning, the answer is generally "yes".
1/
Does an agent need to be "aware" of its executed actions to learn to achieve a certain goal?
In reinforcement learning, the answer is generally "yes".
1/
July 10, 2026 at 1:58 AM
New work led by Filippo out on Neurocomputing! W/ @ksks.bsky.social , @kanair.bsky.social and me.
Does an agent need to be "aware" of its executed actions to learn to achieve a certain goal?
In reinforcement learning, the answer is generally "yes".
1/
Does an agent need to be "aware" of its executed actions to learn to achieve a certain goal?
In reinforcement learning, the answer is generally "yes".
1/
Now online (Neurocomputing), led by Nhut Truong, "Beyond topography: Topographic regularization improves robustness and reshapes representations in convolutional neural networks". Improvements found for image-noise, weight perturbations and adversarial attacks.
www.sciencedirect.com/science/arti...
www.sciencedirect.com/science/arti...
www.sciencedirect.com
July 16, 2026 at 3:18 PM
Now online (Neurocomputing), led by Nhut Truong, "Beyond topography: Topographic regularization improves robustness and reshapes representations in convolutional neural networks". Improvements found for image-noise, weight perturbations and adversarial attacks.
www.sciencedirect.com/science/arti...
www.sciencedirect.com/science/arti...
Biocomputing, neurocomputing: de toekomst van berekeningen maken
https://ioplus.nl/nl/posts/biocomputing-neurocomputing-de-toekomst-van-berekeningen-maken
https://ioplus.nl/nl/posts/biocomputing-neurocomputing-de-toekomst-van-berekeningen-maken
Biocomputing, neurocomputing: de toekomst van berekeningen maken
In het Tech Trends-rapport 2026 schetsen experts een beeld van een aantal nieuwe manieren van computing.
ioplus.nl
December 30, 2025 at 7:05 AM
Biocomputing, neurocomputing: de toekomst van berekeningen maken
https://ioplus.nl/nl/posts/biocomputing-neurocomputing-de-toekomst-van-berekeningen-maken
https://ioplus.nl/nl/posts/biocomputing-neurocomputing-de-toekomst-van-berekeningen-maken
How will your multi-view model perform if data is missing during deployment?
Usually, this translates into a considerable decline in accuracy.
However, simpler approaches in model design can make it robust to missing data. We address this in our recent paper in the Neurocomputing journal.
Usually, this translates into a considerable decline in accuracy.
However, simpler approaches in model design can make it robust to missing data. We address this in our recent paper in the Neurocomputing journal.
April 11, 2025 at 7:21 AM
How will your multi-view model perform if data is missing during deployment?
Usually, this translates into a considerable decline in accuracy.
However, simpler approaches in model design can make it robust to missing data. We address this in our recent paper in the Neurocomputing journal.
Usually, this translates into a considerable decline in accuracy.
However, simpler approaches in model design can make it robust to missing data. We address this in our recent paper in the Neurocomputing journal.
Our paper "Iterated relevance matrix analysis for improved classification and robustness in prototype-based learning schemes" appears in Neurocomputing. This work with first authors Sofie Lövdal and Elina van den Brandhof is
available at
www.sciencedirect.com/science/arti...
🧪 #machinelearning
available at
www.sciencedirect.com/science/arti...
🧪 #machinelearning
Iterated relevance matrix analysis for improved classification and robustness in prototype-based learning schemes
Generalized Matrix Learning Vector Quantization (GMLVQ) is an inherently interpretable classifier, but the learned distance measure may converge into …
www.sciencedirect.com
December 24, 2025 at 10:05 AM
Our paper "Iterated relevance matrix analysis for improved classification and robustness in prototype-based learning schemes" appears in Neurocomputing. This work with first authors Sofie Lövdal and Elina van den Brandhof is
available at
www.sciencedirect.com/science/arti...
🧪 #machinelearning
available at
www.sciencedirect.com/science/arti...
🧪 #machinelearning
Saja's survey on the #intepretability/#explainability of Capsule Networks was accepted at #Neurocomputing. Give it a look while it is free to access.
doi.org/10.1016/j.ne...
#CapsNets #AI #ML #XAI #sqIRL #UAntwerp #IDLab
doi.org/10.1016/j.ne...
#CapsNets #AI #ML #XAI #sqIRL #UAntwerp #IDLab
Redirecting
doi.org
November 25, 2025 at 7:22 AM
Saja's survey on the #intepretability/#explainability of Capsule Networks was accepted at #Neurocomputing. Give it a look while it is free to access.
doi.org/10.1016/j.ne...
#CapsNets #AI #ML #XAI #sqIRL #UAntwerp #IDLab
doi.org/10.1016/j.ne...
#CapsNets #AI #ML #XAI #sqIRL #UAntwerp #IDLab
A model of Frame and Verb Compliance in language acquisition, 2007 Neurocomputing
February 23, 2026 at 11:51 PM
A model of Frame and Verb Compliance in language acquisition, 2007 Neurocomputing
Hermite polynomials facilitating on-line learning analysis of layered neural networks with arbitrary activation function
by O. Citton, F. Richert , M. Biehl, and M. Straat
Neurocomputing Volume 655, 28 November 2025, 131328
Published version available online
www.sciencedirect.com/science/arti...
by O. Citton, F. Richert , M. Biehl, and M. Straat
Neurocomputing Volume 655, 28 November 2025, 131328
Published version available online
www.sciencedirect.com/science/arti...
Hermite polynomials facilitating on-line learning analysis of layered neural networks with arbitrary activation function
Following the standard statistical mechanics methods we analyze the training by online stochastic gradient descent of two-layer neural networks in a s…
www.sciencedirect.com
September 10, 2025 at 5:43 AM
Hermite polynomials facilitating on-line learning analysis of layered neural networks with arbitrary activation function
by O. Citton, F. Richert , M. Biehl, and M. Straat
Neurocomputing Volume 655, 28 November 2025, 131328
Published version available online
www.sciencedirect.com/science/arti...
by O. Citton, F. Richert , M. Biehl, and M. Straat
Neurocomputing Volume 655, 28 November 2025, 131328
Published version available online
www.sciencedirect.com/science/arti...
Our article "Interpretable modelling and visualization of biomedical data" is available online (open access) at Neurocomputing!
www.sciencedirect.com/science/arti...
www.sciencedirect.com/science/arti...
Interpretable modelling and visualization of biomedical data
Applications of interpretable machine learning (ML) techniques on medical datasets facilitate early and fast diagnoses, along with getting deeper insi…
www.sciencedirect.com
February 7, 2025 at 1:56 PM
Our article "Interpretable modelling and visualization of biomedical data" is available online (open access) at Neurocomputing!
www.sciencedirect.com/science/arti...
www.sciencedirect.com/science/arti...
»The aim of PYGMALION was to create an independent European technological base for the applications, algorithms and software aspects of neurocomputing.«
Project end date: February 1st, 1991.
Source: cordis.europa.eu/project/id/2...
#vintagecomputing #ai
Project end date: February 1st, 1991.
Source: cordis.europa.eu/project/id/2...
#vintagecomputing #ai
Neurocomputing | PYGMALION | Projekt | Fact Sheet | FP2 | CORDIS | European Commission
The aim of PYGMALION was to create an independent European technological base for the applications, algorithms and software aspects of neurocomputing. The project led to the coordination of research o...
cordis.europa.eu
July 11, 2025 at 1:15 PM
»The aim of PYGMALION was to create an independent European technological base for the applications, algorithms and software aspects of neurocomputing.«
Project end date: February 1st, 1991.
Source: cordis.europa.eu/project/id/2...
#vintagecomputing #ai
Project end date: February 1st, 1991.
Source: cordis.europa.eu/project/id/2...
#vintagecomputing #ai
I’d recommend Parallel Distributed Processing for good (though older) explanations about the importance of hidden layers mitpress.mit.edu/978026268053...
Parallel Distributed Processing
What makes people smarter than computers? These volumes by a pioneering neurocomputing group suggest that the answer lies in the massively parallel architect...
mitpress.mit.edu
November 5, 2024 at 7:43 PM
I’d recommend Parallel Distributed Processing for good (though older) explanations about the importance of hidden layers mitpress.mit.edu/978026268053...
Maximize your neuroimaging workflows with robust computational environments! Simplify setup, enhance processing speed, and collaborate easier.
🔗 Explore your options: nitrc.org #NeuroComputing #DataScience #ResearchTech #Bioinformatics
🔗 Explore your options: nitrc.org #NeuroComputing #DataScience #ResearchTech #Bioinformatics
November 6, 2025 at 12:45 PM
Maximize your neuroimaging workflows with robust computational environments! Simplify setup, enhance processing speed, and collaborate easier.
🔗 Explore your options: nitrc.org #NeuroComputing #DataScience #ResearchTech #Bioinformatics
🔗 Explore your options: nitrc.org #NeuroComputing #DataScience #ResearchTech #Bioinformatics
I swear, every consequence repeatedly surprising AI companies is covered in the two volumes of Parallel Distributed Processing. mitpress.mit.edu/978026268053...
Parallel Distributed Processing
What makes people smarter than computers? These volumes by a pioneering neurocomputing group suggest that the answer lies in the massively parallel architect...
mitpress.mit.edu
June 9, 2024 at 2:14 AM
I swear, every consequence repeatedly surprising AI companies is covered in the two volumes of Parallel Distributed Processing. mitpress.mit.edu/978026268053...
Our work on Interpretable #Hyperdimensional Computing (HDC) classifiers for #tabular data is now available at #Neurocomputing.
doi.org/10.1016/j.ne...
#Neuromorphic #interpretability #Explainability #VSA #xai #ML #FAIR #UAntwerp #imec
doi.org/10.1016/j.ne...
#Neuromorphic #interpretability #Explainability #VSA #xai #ML #FAIR #UAntwerp #imec
Redirecting
doi.org
November 4, 2025 at 11:46 AM
Our work on Interpretable #Hyperdimensional Computing (HDC) classifiers for #tabular data is now available at #Neurocomputing.
doi.org/10.1016/j.ne...
#Neuromorphic #interpretability #Explainability #VSA #xai #ML #FAIR #UAntwerp #imec
doi.org/10.1016/j.ne...
#Neuromorphic #interpretability #Explainability #VSA #xai #ML #FAIR #UAntwerp #imec
Nuevo artículo publicado en acceso abierto:
Comparative analysis of unsupervised anomaly detection techniques for heat detection in dairy cattle
#Neurocomputing #anomalydetection #smartcollars #dairycattle
doi.org/10.1016/j.ne...
Comparative analysis of unsupervised anomaly detection techniques for heat detection in dairy cattle
#Neurocomputing #anomalydetection #smartcollars #dairycattle
doi.org/10.1016/j.ne...
December 16, 2024 at 7:00 AM
Nuevo artículo publicado en acceso abierto:
Comparative analysis of unsupervised anomaly detection techniques for heat detection in dairy cattle
#Neurocomputing #anomalydetection #smartcollars #dairycattle
doi.org/10.1016/j.ne...
Comparative analysis of unsupervised anomaly detection techniques for heat detection in dairy cattle
#Neurocomputing #anomalydetection #smartcollars #dairycattle
doi.org/10.1016/j.ne...
The importance of lunch breaks cannot be overrated! This work emerged from a discussion at ESANN2024, was presented at ESANN2025 and extended for Neurocomputing.
With @hsmw.bsky.social colleagues, @bernoulliinstitute.bsky.social : S Lövdal, EL van den Brandhof and M Biehl
doi.org/10.1016/j.ne...
With @hsmw.bsky.social colleagues, @bernoulliinstitute.bsky.social : S Lövdal, EL van den Brandhof and M Biehl
doi.org/10.1016/j.ne...
Redirecting
doi.org
March 8, 2026 at 12:37 PM
The importance of lunch breaks cannot be overrated! This work emerged from a discussion at ESANN2024, was presented at ESANN2025 and extended for Neurocomputing.
With @hsmw.bsky.social colleagues, @bernoulliinstitute.bsky.social : S Lövdal, EL van den Brandhof and M Biehl
doi.org/10.1016/j.ne...
With @hsmw.bsky.social colleagues, @bernoulliinstitute.bsky.social : S Lövdal, EL van den Brandhof and M Biehl
doi.org/10.1016/j.ne...
New unit paper on “Decoding pre-movement neural activity from thalamic LFPs for adaptive neurostimulation in tremor patients”, published in Neurocomputing.
mrcbndu.ox.ac.uk/papers/decod...
@mrcbndu.ox.ac.uk @ndcnoxford.bsky.social @medsci.ox.ac.uk @ox.ac.uk
mrcbndu.ox.ac.uk/papers/decod...
@mrcbndu.ox.ac.uk @ndcnoxford.bsky.social @medsci.ox.ac.uk @ox.ac.uk
February 10, 2026 at 4:22 PM
New unit paper on “Decoding pre-movement neural activity from thalamic LFPs for adaptive neurostimulation in tremor patients”, published in Neurocomputing.
mrcbndu.ox.ac.uk/papers/decod...
@mrcbndu.ox.ac.uk @ndcnoxford.bsky.social @medsci.ox.ac.uk @ox.ac.uk
mrcbndu.ox.ac.uk/papers/decod...
@mrcbndu.ox.ac.uk @ndcnoxford.bsky.social @medsci.ox.ac.uk @ox.ac.uk
New from the Urban Lab: PyfUS!
An open-source Python toolkit for analyzing whole-brain fUS data—now with single-voxel clustering for fine-scale insights. Just published in Neurocomputing.
🔗 Paper: https://doi.org/10.1016/j.neucom.2025.130899
🛠️ Code: github.com/OpenfUS/PyfUS
An open-source Python toolkit for analyzing whole-brain fUS data—now with single-voxel clustering for fine-scale insights. Just published in Neurocomputing.
🔗 Paper: https://doi.org/10.1016/j.neucom.2025.130899
🛠️ Code: github.com/OpenfUS/PyfUS
July 21, 2025 at 6:37 AM
New from the Urban Lab: PyfUS!
An open-source Python toolkit for analyzing whole-brain fUS data—now with single-voxel clustering for fine-scale insights. Just published in Neurocomputing.
🔗 Paper: https://doi.org/10.1016/j.neucom.2025.130899
🛠️ Code: github.com/OpenfUS/PyfUS
An open-source Python toolkit for analyzing whole-brain fUS data—now with single-voxel clustering for fine-scale insights. Just published in Neurocomputing.
🔗 Paper: https://doi.org/10.1016/j.neucom.2025.130899
🛠️ Code: github.com/OpenfUS/PyfUS
Our recent work, validated with heterogenous real-world data from the Earth observation domain, can be accessed (open-access 🔓) at the Neurocomputing journal
www.sciencedirect.com/science/arti...
www.sciencedirect.com/science/arti...
Missing data as augmentation in the Earth Observation domain: A multi-view learning approach
Multi-view learning (MVL) leverages multiple sources or views of data to enhance machine learning model performance and robustness. This approach has …
www.sciencedirect.com
April 11, 2025 at 7:21 AM
Our recent work, validated with heterogenous real-world data from the Earth observation domain, can be accessed (open-access 🔓) at the Neurocomputing journal
www.sciencedirect.com/science/arti...
www.sciencedirect.com/science/arti...