With just four molecules, you get a grid cell attractor with functional 2D path integration. With N molecules, you get exponentially many Hebbian assemblies with discrete attractor dynamics. With smart biases between the molecules, you can get any dynamical system. Sequences, motor programs, …

Dendrites!

New preprint with @sevberg.bsky.social! We map Hopfield-like binary networks onto spiking networks with dendrites … and it works! Same memory capacity, bigger basins of attraction, plus selective recall through dendritic gating, and more. How? Dendrites! See below.

*** New paper from the lab in Nature Comms !! *** “Noradrenaline causes a spread of association in the hippocampal cognitive map” by the amazing @reneecabbage.bsky.social in collab. w labs of Jill O’Reilly, Will Clarke, and modeling from Prakriti Parthasarathy and @tpvogels.bsky.social, see below!

Most of us are familiar with the nightingale's song, but did you know there's more to it than just a melody? Read more: doi.org/10.1016/j.cu... @danielavallentin.bsky.social @tpvogels.bsky.social @mpiforbi.bsky.social ​

This paper shows alignment between LLMs and brain data is outperformed by a null model. More evidence for the argument I've been making in talks lately that we shouldn't believe any computational paper that puts less effort into null than main model. www.biorxiv.org/content/10.1...

This is such a cool idea! Not long term "networks" of young profs that create and cement in in-groups and out-groups, but summer schools, tailored to the acute needs of the young PI. Transient, but tailored, tidal and totally timeous, too. Terrific. Try today!

...in which @danielavallentin.bsky.social shows: vogels match rivals’ whistles, imitating pitch + duration. W'sm, a vogelsmodel model reveals hierarchical organisation of vogelsongs: Duration constrains pitch, a trade-off. Not all vogels, just nightingales. www.sciencedirect.com/science/arti...

Deadline for poster submissions in Jan 27!! That totally escaped me!

Excited to see the paper fully published. It's an important milestone for training SNNs with exact gradients, replacing our earlier tricks of a "delay line augmentation" to capture temporal relationships. Delays can now be learnt alongside weights naturally. Amazing work @mbalazs98.bsky.social !

13 yrs ago some crusty prof asked me if I were "able to teach quantum chromo dynamics" during a faculty interview, as a means to disqualify me (his preferred candidate got the job, surprise!). These days, I am subjugated vicariously, when my postdocs experience the same BS. WTAF. So infuriating.

Last of my fave books back on a shelf, after 5 years in a box. I wouldn’t have thought it takes a full grant cycle to move but here we are. And there’s still more boxes. Of garbage. Mostly. But could not be happier to have the books back

"Spiking Networks Hate It! Find Out the One Plasticity Trick They Don’t Want You to Know! Never stabilise models by hand again." - I woke up thinking we missed an opportunity with the title of this one. :/ www.science.org/doi/10.1126/... Also: It snowed in Vienna, 10cm white fluffies! Happy Sunday!

...and when I say @worldwideneuro.bsky.social, I really mean @bozelosp.bsky.social, obvs. Also, ...there is an all new front end for worldwideneuro coming soon. Stay tuned!

Here is the product of a collab between @thetransmitter.bsky.social, @worldwideneuro.bsky.social & @moritzstefaner.bsky.social A portrait of the #StateofNeuro 2025: Text mining of (publicly available) neuro resources: Easy in concept, tricky in realisation, made possible by @simonsfoundation.org 🙏

I just found this stub in a jacket that I guess I have not worn in twenty three years? 12 bucks for Lali Puna, sweat and spit and all, at the (tiny) Middle East downstairs on Mass Ave. Those were the days. So so good.

Maybe check out the associated website. icg-explorer.org. I know I know, it's not fully integrated with ICG1, but these things are difficult.

Congratulations Juan! Exciting news!

Some addendum here. The '17 elife paper only has 43 citations in 8 years :( but the associated ionchannelmodels.org had 900+ unique visitors in 2024. Some 180 of them used the "compare ion channel models" function, pretty deep into the page. Every two days someone used a service we didn't advertise!

What Americans Die from? And what the media reports on... Our World in Data ourworldindata.org/does-the-new...

The working title of this project was “super models”, but we chickened out in carrying that terminology through. We were worried anyone was gonna think we were making fun of human supermodels, or worse, ion channels!! And we weren’t sure blue steel was the right association for a CA++ channel…

“Mapping ion channel function” doi.org/10.7554/eLif... isn’t exactly a citation slayer, but it’s still one of my favourites (& my first independent project). Today we push pt 2, where we trace code origin & unite almost all channel models in a common expression. Boom! www.biorxiv.org/content/10.1...

No more Paris-Wien/Berlin night trains "Die ÖBB bedauern, dass nach dem Rückzug der französischen Partner, die beiden Nachtzugverbindungen ab 14. Dez 2025 nicht mehr angeboten werden können." - SNCF, this is you presse-oebb.at/news-oebb-be... My analysis in English: jonworth.eu/sncfs-half-h...

"How many disco balls have you purchased in your life time?"

Neuronal diversity is written in transcriptional codes 🧬. But what is the logic of these codes that define cell types and wiring patterns? To find out we built a #scRNAseq developmental atlas of the Drosophila nerve cord and linked it to the #connectome 🪰🧠 #preprint thread ⬇️1/8

Involuntary memory retrieval is a common phenomenon in autobiographical #memory, and a symptom of #PTSD. This study reveals the neural mechanism of #InvoluntaryMemory, showing that involuntary & voluntary memories rely on distinct neural processes @plosbiology.org 🧪 plos.io/4oQkb3e

Almost every major university financial fraud is done by administrators...yet note the difference between the comically instrusive surveillance and accountability for faculty over every dollar vs the complete lack of oversight for management spending (or stealing). lfpress.com/news/local-n...

I now had a chance to listen to it and it’s not entirely terrible. Gaute’s voice is actually quite soothing and his questions are good. You might not immediately die while listening to this. Thanks @gauteeinevoll.bsky.social for cutting out all the bits that would’ve gotten me fired.

Episode #30 in #TheoreticalNeurosciencePodcast: On co-dependent excitatory and inhibitory plasticity – with Tim Vogels @tpvogels.bsky.social theoreticalneuroscience.no/thn30 How can excitatory and inhibitory synaptic plasticity, co-exist in spiking neural network models? With co-dependence!

then lettuce rejoice where we at

Arrived in Okinawa last night for teaching at the #OIST summer school. So far so good, almost as pretty as the #imbizo ;) Looking forward to a week of science!

Here is your last reminder that the application deadline for Imbizo.Africa is nearing quickly, the 1st of July, in fact tomorrow. Still the place where diversity is at its best in the world! Tell all who need to hear. #africa #neuro

Hey @biorxiv-neursci.bsky.social your verification of my humanity is stuck in a forever loop, and it kinda ..... if my frustration is any measure of my humanity then I am defos HUMAN. Ayee!

We just pushed “Memory by a 1000 rules” onto bioRxiv, where we use clever #ML to find #plasticity quadruplets (EE, EI, IE, II) that learn basic stability in spiking nets. Why is it cool? We find 1000s!! of solutions, and they don’t just stabilise. They #memorise! www.biorxiv.org/content/10.1...

What if there were grants for which a funder phrases a (well constrained) question to be answered, and the cash they will pay for the answer, and P.I.'s lay out a concrete action plan + a narrative track record that argues their appropriate expertise? They can also ask for more cash, if necessary.

This is just a quick reminder, www.youtube.com/watch?v=QFUz...

Last month, we organized #ODIN2025 @alleninstitute.org (along w Karel, Saskia @sejdevries.bsky.social , David & Satra) with the theme "Integrating Scales & Modalities". See the videos📣👇 Full playlist: youtube.com/playlist?lis... 🎥 Speaker: alleninstitute.org/odin-symposi... youtu.be/N1IMki0RnIc?...

Doing Neuroscience in ….

87 years ago. In April 1938, Guy Callendar published the first evidence from thermometer observations that the world’s land areas were warming. He also linked the observed warming to the increase in atmospheric CO₂ from burning coal. 87 years ago. rmets.onlinelibrary.wiley.com/doi/abs/10.1...

… and tiredness

Not _exactly_ my usual thing, but here it is, in full Dolby Surround, w/ Sebas Calderon Garcia, Giacomo Costalunga, & @danielavallentin.bsky.social : “Nightingales imitate the duration of whistle syllables” [not just the *pitch*, & concurrent w/ ~ 🐦‍⬛]. Next, how? www.biorxiv.org/content/10.1...

www.nytimes.com/2025/04/03/o... Really important piece to share widely. Too many Americans don’t know how severe a threat this Administration poses to our scientific infrastructure, global leadership, and health security

About time, @istaresearch.bsky.social welcome !

Happy to share the recording of my plenary talk at Cosyne 2025 two days ago. You will learn about the statistical physics approach, phase transitions in learning, transformers, sequence models, attention etc. www.youtube.com/watch?v=PurZ...

Leaving Vienna in the best way how. #COSYNE2025 here we come, ready or not.

Roses are red Violets are blue

"Evidence suggests the local recurrent network encodes the structure of natural sensory input. & that it does so via active filtering, transforming network inputs to boost those associated w/ nat'l sensation" Amidst <waves hands> all this, I have a new review article that... arxiv.org/abs/2501.1...

In my city there is one cybertruck and one DIY parody of a cybertruck and the parody truck has become a local celebrity. Meanwhile, I always wish I had some rotten eggs to throw at the tesla whenever I see it.