I just discovered the ARC-AGI initiative and the associated test to estimate how close "AI" models are from #AGI

https://arcprize.org/arc-agi

While I found the initiative interesting, I'm not sure I understand what in this test really guarantees that the model is capable of some form of generalization and problem-solving.
Wouldn't it be possible for specialized pattern-matching/discovering algorithms to solve such problems?
I imagine some computer scientists, mathematicians or computational neuroscientists have already had a look at this, so would anyone knows of some articles/blogs on the topic?

Maybe @wim_v12e? Is this something you already looked at?

A new version of the Open Source Brain (#OpenSourceBrain) model validation framework was just released. Please update your installations:

```
pip install -U OSBModelValidation
```

https://github.com/OpenSourceBrain/osb-model-validation/releases/tag/v0.3.9

#OMV is a #Python package that allows you to validate your #NeuroML models against different simulation engines---to ensure that you get the same behaviours on all these engines. Examples:

https://github.com/OpenSourceBrain/.github/blob/main/testsheet/README.md

A new release of PyNeuroML is available, please update your installations:

https://github.com/NeuroML/pyNeuroML/releases

```
pip install --upgrade pyneuroml
```

Almost last call to register for UK neural computation conference in London July 10-11. Registration deadline is July 1st. We have some great talks and posters as well as a session on funding with ARIA.

Look forward to seeing you all there. Now click here

https://neuralcomputation.uk/

A new release of #PyNeuroML is available. Please update to get the latest fixes and features. #NeuroML #ComputationalModelling #ComputationalNeuroscience

```
pip install --upgrade pyneuroml
```

If you're seeing your #NEST simulator builds fail recently, it's because of a new #Cython release:

https://github.com/nest/nest-simulator/issues/3495

Workaround: fix `cython<3.1.0` for the moment (until new NEST releases with the fix are made)

New #preprint! Confavreux et al. use meta-learning to uncover thousands of diverse, local #plasticity rule quadruplets that stabilize #RecurrentSpikingNetworks — and incidentally support #memory functions like novelty detection, replay, & contextual prediction. A striking case of function emerging from stability.

https://doi.org/10.1101/2025.05.28.656584

The #Italian Network of #ComputationalNeuroscience announced its 2025 conference:

Palazzo della Salute, Padova, Italy
September 22–24, 2025
Submission deadline: June 7, 2025
https://www.incn.it/

A 3-day deep dive into the brain — from models to data, theory to technology.
#CompNeuro #Neuroscience

In the next version, 0.6.6, all component classes in the #libNeuroML #Python API gain a quick method to list the parameters of their model objects. So, if you load a #NeuroML file/model, you can easily print the parameters of the different model entities. See the test in the pull request for a quick example:

https://github.com/NeuralEnsemble/libNeuroML/pull/208/files

Researchers in France are working on creating a French network of researchers to organize interaction, communication and training in #computationalneuroscience

If you are a CompNeuro working in France, consider joining, and registering to our mailing list: https://listes.services.cnrs.fr/wws/subscribe/rt_neurocomp

https://bsky.app/profile/lauradugue.bsky.social/post/3lo6rrtvb3k2v
https://www.linkedin.com/posts/laura-dugu%C3%A9-59964756_rtneurocomp-r%C3%A9seau-fran%C3%A7ais-de-neurosciences-activity-7324039702427090944-LKsN

Nouveau Réseau Français De Neurosciences Computationnelles !

Une initiative passionnante est en cours en France ! Des chercheurs établissent un réseau national dédié aux #NeurosciencesComputationnelles pour faciliter :

• La collaboration entre équipes de recherche
• Le partage de connaissances et de ressources
• La formation de la prochaine génération de chercheurs
• La visibilité de la recherche française dans ce domaine en pleine expansion

Vous travaillez en #ComputationalNeuroscience sur le territoire français ? Que vous soyez chercheur·e confirmé·e, post-doctorant·e, doctorant·e ou étudiant·e, rejoignez notre communauté !

Inscrivez-vous à notre liste de diffusion pour recevoir les actualités, événements et opportunités :

https://listes.services.cnrs.fr/wws/subscribe/rt_neurocomp

Ensemble, renforçons l'excellence française en #NeurosciencesComputationnelles !
#Recherche #Neuroscience #CNRS #Science #IA #IntelligenceArtificielle

A few weeks ago, I shared a differential equations tutorial for beginners, written from the perspective of a neuroscientist who's had to grapple with the computational part. Following up on that, I've now tackled the first real beast encountered by most computational neuroscience students: the Hodgkin-Huxley model.

While remaining incredibly elegant to this day, this model is also a mathematically dense system of equations that can overwhelm and discourage beginners, especially those with non-mathematical backgrounds. Similar to the first tutorial, I've tried to build intuition step-by-step, starting with a simple RC circuit, layering in Na⁺ and K⁺ channels, and ending with the full spike-generation story.

Feedback is welcome, especially from fellow non-math converts.
https://neurofrontiers.blog/building-a-virtual-neuron-2/

#ComputationalNeuroscience #Python #hodgkinHuxleyModel #math #biophysics

From: @neurofrontiers
https://neuromatch.social/@neurofrontiers/114150087773204813

How do babies and blind people learn to localise sound without labelled data? We propose that innate mechanisms can provide coarse-grained error signals to boostrap learning.

New preprint from @yang_chu.

https://arxiv.org/abs/2001.10605

Thread below

Preview of the talk I'm giving on Friday. #neuroscience #CompNeuro #ComputationalNeuroscience

When I transitioned from cognitive to computational neuroscience, I found myself in a bit of a bind. I had learned calculus, but I had progressed little beyond pattern recognition: I knew which rules to apply to find solutions to which equations, but the equations themselves lacked any sort of real meaning for me.

So I struggled with understanding how formulas could be implemented in code and why the code I was reading could be described by those formulas. Resources explaining math “for neuroscientists” were unfortunately quite useless for me, because they usually presented the necessary equations for describing various neural systems, assuming the presence of that basic understanding/intuition I lacked.

Of course, I figured things out eventually (otherwise I wouldn’t be writing about it), but I’m 85% sure I’m not the only one who’s ever struggled with this, and so I wrote the tutorial I wish I could’ve had. If you’re in a similar position, I hope you’ll find it useful. And if not, maybe it helps you get a glimpse into the struggles of the non-math people in your life. Either way, it has cats.

https://neurofrontiers.blog/building-a-virtual-neuron-1/

#NeuroML is participating in #GSoC2025 again this year under @INCF . We're looking for people with some experience of #ComputationalNeuroscience to work on developing #standardised biophysically detailed computational models using #NeuroML #PyNN and #OpenSourceBrain.

Please spread the word, especially to students interested in modelling. We will help them learn the NeuroML ecosystem so they can use its standardised pipeline in their work.

https://docs.neuroml.org/NeuroMLOrg/OutreachTraining.html#google-summer-of-code

CC #AcademicChatter

I'm giving an online talk starting in 15m (as part of UCL's NeuroAI series).

It's on neural architectures and our current line of research trying to figure out what they might be good for (including some philosophy: what might an answer to this question even look like?).

Sign up (free) at this link to get the zoom link:

https://www.eventbrite.co.uk/e/ucl-neuroai-talk-series-tickets-1189972031379?utm-campaign=social&utm-content=attendeeshare&utm-medium=discovery&utm-term=listing&utm-source=cp&aff=ebdsshcopyurl

With the current situation in the #US, several of my former colleagues there are looking for a #PostDocJob in #Europe, to do #BehaviouralNeuroscience or #ComputationalNeuroscience in #SpatialCognition (or adjacent).
Lots of hashtags I know..

Do you know a #EU or #UK #Neuroscience lab looking to hire a postdoc in these fields? Let me know and I'll pass it on to them!

Edit: adding #RodentResearch and #humanresearch for the species concerned (in this case)

Come along to my (free, online) UCL NeuroAI talk next week on neural architectures. What are they good for? All will finally be revealed and you'll never have to think about that question again afterwards. Yep. Definitely that.

Wed 12 Feb 2025
2-3pm GMT
Details and registration: https://www.eventbrite.co.uk/e/ucl-neuroai-talk-series-tickets-1216638381149

Neuromatch Academy 2025 is coming!

Key Dates:
Feb 24 – Applications Open!
Mar 23 – Deadline (Midnight in your timezone)
Mid-April – Decisions Announced
Early May – Enrollment Deadline

Join a global community & dive into Computational Neuroscience, Deep Learning, Comp Tools for Climate Science, or NeuroAI! Don’t miss out—apply & tag a friend!