Eilif B. Muller

Congrats @clayreid.bsky.social @andreastolias.bsky.social @sebastianseung.bsky.social and the MICrONs team!
Glad to see this heroic dataset is finally out!
www.nature.com/immersive/d4...
#neuroskyence #neuroai

Attempting the impossible: A 20-year journey to learn the language of the brain In what is considered the most complicated neuroscience experiment ever attempted, scientists from the Allen Institute and global collaborators have created the largest, most complex wiring diagram an...

I'm a big fan of the MICrONs project. It's a huge step forward for understanding cortical structure and function, and a model of open science. This is an amazing artful website that communicates why!
alleninstitute.org/news/learnin...

We are excited to announce the World premiere of HOMMAGE! Nominated for Best Art World & Best Art Experience at Raindance Immersive 2025 • Tour reservations open on 2nd June,
www.raindanceimmersive.com#/2025-hommage/
#Moebius #Resonite

Learning to combine top-down context and feed-forward representations under ambiguity with apical and basal dendrites One of the hallmark features of neocortical anatomy is the presence of extensive top-down projections into primary sensory areas, with many impinging on the distal apical dendrites of pyramidal neuron...

here some recent modeling work in this direction:
arxiv.org/abs/2312.05484
elifesciences.org/reviewed-pre...
elifesciences.org/reviewed-pre...
www.biorxiv.org/content/10.1...

I would prefer to ask the question thus: What does PFC learn from V1 to IT, and vice versa. I think we'll be better equipped to answer these questions once we understand learning in cortical hierarchies better ...

It's circular though, since what a cortical region "does" is to a important degree determined by the inputs it receives from other regions ... thus I find "read-out" to be not quite the right word for how a higher-order cortical region may be influenced by regions closer to the sensory periphery.

In vague terms, my guess: if a region is useful for another region at the "right time" during development/and or evolution, then a white matter connection would exist/persist in that resp. direction, with its anatomical strength proportional to its usefulness. Same for other direction.

Harris et al., 2019 has some data on this in mouse ... they found reciprocal direct connections between VISp and ACAd (Ext. Data Fig 8; www.nature.com/articles/s41...) and resolved their laminar profiles

Distinct roles of PV and Sst interneurons in visually induced gamma oscillations Gamma-frequency oscillations are a hallmark of active information processing and are generated by interactions between excitatory and inhibitory neuro…

Happy to see this study led by Irene Onorato finally out - we show distinct phase locking and spike timing of optotagged PV cells and Sst interneuron subtypes during gamma oscillations in mouse visual cortex, suggesting an update to the classic PING model www.sciencedirect.com/science/arti...

Dendritic arbors structure memories Synapses on different dendritic domains store distinct types of information

A new Science study finds that different dendritic segments of a single neuron follow distinct rules.

The results challenge the idea that neurons follow a single learning strategy and offer a new perspective on how the brain learns and adapts behavior. scim.ag/3YQdUZK #SciencePerspective

Hippocampal ripples initiate cortical dimensionality expansion for memory retrieval How are past experiences reconstructed from memory? Learning is thought to compress external inputs into low-dimensional hippocampal representations, later expanded into high-dimensional cortical acti...

🔎 In sum: Hippocampal ripples trigger a transition from compressed hippocampal codes to expanded cortical states that reconstruct past experiences.

Great collaboration with @s-michelmann.bsky.social and @doellerlab.bsky.social!

Read the full paper here:
www.biorxiv.org/content/10.1...

Check out our new paper!
Vision models often struggle with learning both transformation-invariant and -equivariant representations at the same time.

@hafezghm.bsky.social shows that self-supervised prediction with proper inductive biases achieves both simultaneously. (1/4)
#MLSky #NeuroAI

This is bleak reading

For understanding learning in neocortex 🧠, self-supervised learning 🤖 is interesting, but has several shortcomings.

Seq-JEPA is a step in the right direction. It learns by predicting sensory outcomes from series of interactions. Cool things emerged! 👇with @shahabbakht.bsky.social
#MLSky #NeuroAI

Top-down feedback is ubiquitous in the brain and computationally distinct, but rarely modeled in deep neural networks. What happens when a DNN has biologically-inspired top-down feedback? 🧠📈

Our new paper explores this: elifesciences.org/reviewed-pre...

I'm new to @resonite.com , but I think its very important what you're doing. It was your eye tracking in NeOS yt videos that really got my attention. You're a leader. Followers fit in. Leaders are busy taking first steps. You have so many followers, me included, finding how we fit in your world.

I would also appreciate if you could give some hints what's getting you from the photos to the mesh you import in resonite?

After a great conference in Boston, CCN is going to take place in Amsterdam in 2025! To help the exchange of ideas between #neuroscience, cognitive science, and #AI, CCN will for the first time have full length paper submissions (alongside the established 2 pagers)! Info below👇
#NeuroAI #CompNeuro

2/2 But we didn't stop there: In a 3rd paper (also @elife.bsky.social , we simulated neocortical reorganization due to invivo like input and calcium-based plasticity mechanisms ... our predictions for emerging motifs were confirmed in the MICrONS EM connectome
doi.org/10.7554/eLif...

1/2 Migrating to Bsky, and following @elife.bsky.social I notice our 2 big papers on simulation of neocortical circuits are at the top of the feed!

How to migrate from X to Bluesky without losing your followers This extension lets you easily migrate your follows and block list from X (formerly Twitter) to Bluesky, but you need to act fast because its functionality may be short-lived. Here's why.

Have been off twitter for over a year, tried mastadon ... Bsky looks more like its in the goldilocks zone. Migrated my network using "Sky Follower Bridge"
www.zdnet.com/article/how-...
or is there a better approach? I recall a "NeuroMigration" movement for mastadon ...

Add me please 🙋

Sequential predictive learning is a unifying theory for hippocampal representation and replay bioRxiv - the preprint server for biology, operated by Cold Spring Harbor Laboratory, a research and educational institution

A little late on the 🦋🚂, but happy to share a new(ish) preprint - the culmination of three years of postdoc with @tyrellturing.bsky.social and @apeyrache.bsky.social, and my first real foray into #NeuroAI as a tool to study the sleeping brain 🧠🟦:

biorxiv.org/content/10.1...

(1/🧵) c’est parti!