Interested in multitasking RNNs from a theory perspective? Check out this preprint with the amazing @lpezon.bsky.social. TLDR: A low-rank / latent space approach to @lndriscoll.bsky.social & @sussillodavid.bsky.social style shared components. And a story starring solution space degeneracy.

Louis Pezon@lpezon.bsky.social · 2mo ago

Both brains and RNNs can re-use components of computation across similar tasks or contexts. But what exactly are those “shared components”? How can they be used to solve several tasks? We address these questions in a new preprint with @avm.bsky.social! Link: www.biorxiv.org/content/10.6...

How can RNNs combine independently learned computations without being trained on their combinations? Excited to share our new preprint📜 www.biorxiv.org/content/10.6...

There seems to be a broad perception across psychology and neuroscience that work shouldn't be "too technical" in order to reach the broadest possible audience. While I think we should strive for accessibility, I feel that this attitude can also be self-defeating: why are we dumbing down?

Come learn some programming with a healthy helping of critical thinking; something no genAI can take away from you :) Also Prague is great fun ;)

Anna-Lena Eckert@eckertal.bsky.social · 4mo ago

Sadly, applications to the Advanced Python summer school have dropped significantly over the past 2 years. Plus, there'll be no external funding for the 1st time in *17 years*. Likely all because of GenAI - but programming skills still matter🔥 Deadline May 3, please help by sharing: aspp.school

Our latest publication grapples with how the brain could implement gradient descent by sending learning targets top-down, gating plasticity with dendritic inhibition, and updating synaptic weights with biologically observed learning rules like BTSP. www.cell.com/cell-reports...

Cellular and subcellular specialization enables biology-constrained deep learning

Galloni et al. introduce “dendritic target propagation”: a Dale’s law-compliant learning algorithm for cortical microcircuits with soma- and dendrite-targeting inhibition and realistic connectivity co...

cell.com

Aaron Milstein@neurosutras.bsky.social · last yr.

New #NeuroAI #compneurosky preprint! To better understand how target-directed learning works in the brain, we sought to engineer an artificial neural network capable of solving complex image classification tasks that comprises only experimentally-supported biological building blocks. (1/15)

If you are into networks, development, and dynamics don't miss my poster [3-072] this afternoon at #cosyne2026 🤠🧠

I tracked every keyword in 22 years of Cosyne abstracts to map how computational neuroscience evolved — from Bayesian brains to neural manifolds to LLMs — and where it's heading next.

22 years of Brain Science: what CoSyNe tells us about the evolution of Neuroscience

Tracking the intellectual DNA of Computational and Systems Neuroscience through its flagship meeting

open.substack.com

This paper on how the brain may do gradient descent is very cool: www.nature.com/articles/s41...

Vectorized instructive signals in cortical dendrites - Nature

Mice learning a neurofeedback brain–computer interface task show neuron-specific teaching signals in cortical dendrites, consistent with a vectorized solution for credit assignment in the brain.

nature.com

How can RNNs learn continuously without forgetting? 🧠 Our new preprint shows how a predictive learning rule organizes recurrent dynamics into orthogonal manifolds, reducing task interference. Congrats Zihan @zihan-liu.bsky.social ! www.biorxiv.org/content/10.6...

🧪🧠 New preprint: helping resolve a decades-long debate in synaptic plasticity NMDA receptors are central to Hebbian learning. Yet for >30 years, the existence and function of presynaptic NMDA receptors have remained controversial. 📄 doi.org/10.64898/202... 1/6

Learning sculpts orthogonal task manifolds for continual skill learning in recurrent networks https://www.biorxiv.org/content/10.64898/2026.02.15.705283v1

Zajzon, Bouhadjar, Fabre, Schmidt, Ostendorf, Neftci, Morrison, Duarte: SymSeqBench: a unified framework for the generation and analysis of rule-based symbolic sequences and datasets https://arxiv.org/abs/2512.24977 https://arxiv.org/pdf/2512.24977 https://arxiv.org/html/2512.24977

Retinotopy constrains the topology of neural manifolds in macaque visual cortex https://www.biorxiv.org/content/10.64898/2025.12.22.695909v1

This report in Nature on the costs of competing for & administering scientific grants is shocking: "In other words, European taxpayers will have spent more on the funding process than on the funding itself, and the scientific ecosystem has been drained." www.nature.com/articles/d41... 🧪

Point of no returns: researchers are crossing a threshold in the fight for funding

With so little money to go round, the costs of competing for grants can exceed what the grants are worth. When that happens, nobody wins.

nature.com

SAVE THE DATE! On June 16-17, 2026, the NEST Initiative is excited to invite everyone interested in Neural Simulation Technology and the NEST Simulator to the NEST virtual Conference 2026. Registration and abstract submission will open soon! nest-simulator.org/conference

0/10 Thanks for the interest in our preprint. Some takes say it negates or fully supports the “manifold hypothesis”, neither quite right. Our results show that if you only focus on the manifold capturing most of task-related variance, you could miss important dynamics that actually drive behavior.

Dan Levenstein@dlevenstein.bsky.social · 9mo ago

“Our findings challenge the conventional focus on low-dimensional coding subspaces as a sufficient framework for understanding neural computations, demonstrating that dimensions previously considered task-irrelevant and accounting for little variance can have a critical role in driving behavior.”

Y’all are reading this paper in the wrong way. We love to trash dominant hypothesis, but we need to look for evidence against the manifold hypothesis elsewhere: This elegant work doesn't show neural dynamics are high D, nor that we should stop using PCA It’s quite the opposite! (thread)

Dan Levenstein@dlevenstein.bsky.social · 9mo ago

“Our findings challenge the conventional focus on low-dimensional coding subspaces as a sufficient framework for understanding neural computations, demonstrating that dimensions previously considered task-irrelevant and accounting for little variance can have a critical role in driving behavior.”

BTSP in V1 www.biorxiv.org/content/10.1...

Plateau potentials are instructive signals for behavioral timescale synaptic plasticity in the neocortex

Learning occurs via the adjustment of synaptic weights across a variety of timescales. The mechanisms supporting these processes, from single-shot to iterative learning, are unclear. The prevailing mo...

biorxiv.org

A study led by Cina Aghamohammadi is now out in ‪@natcomms.nature.com‬! We developed a mathematical framework for partitioning spiking variability, which revealed that spiking irregularity is nearly invariant for each neuron and decreases along the cortical hierarchy. www.nature.com/articles/s41...

Undecided which workshop to attend at #BernsteinConference ? We put together a stellar lineup for a satellite workshop on the effects of top-down signals on neuronal dynamics! 🧠✨📣 Join us tomorrow in room 0.101 ! bernstein-network.de/bernstein-co... With @ackurth.bsky.social

On the first workshop talk (Monday), I'll make the case that everything might be everywhere when you decode, but that is not the end of the story: different regions have very different dynamics and encoding geometries. This workshop is organized by @aitormg.bsky.social and @ackurth.bsky.social

Jonas Oberste-Frielinghaus, Anno C. Kurth, Julian G\"oltz, Laura Kriener, Junji Ito, Mihai A. Petrovici, Sonja Gr\"un: Synchronization and semantization in deep spiking networks https://arxiv.org/abs/2508.12975 https://arxiv.org/pdf/2508.12975 https://arxiv.org/html/2508.12975

Jakob Stubenrauch, Naomi Auer, Richard Kempter, Benjamin Lindner: Stochastic synaptic dynamics under learning https://arxiv.org/abs/2508.13846 https://arxiv.org/pdf/2508.13846 https://arxiv.org/html/2508.13846

Our current approach to defining neural populations is largely arbitrary. We need new methods for grouping cells, ideally by their dynamics, writes @markdhumphries.bsky.social #neuroskyence www.thetransmitter.org/systems-neur...

The challenge of defining a neural population

Our current approach is largely arbitrary. We need new methods for grouping cells, ideally by their dynamics.

thetransmitter.org

1/3) This may be a very important paper, it suggests that there are no prediction error encoding neurons in sensory areas of cortex: www.biorxiv.org/content/10.1... I personally am a big fan of the idea that cortical regions (allo and neo) are doing sequence prediction. But... 🧠📈 🧪

Sensory responses of visual cortical neurons are not prediction errors

Predictive coding is theorized to be a ubiquitous cortical process to explain sensory responses. It asserts that the brain continuously predicts sensory information and imposes those predictions on lo...

biorxiv.org

Undecided whether to attend the #BernsteinConference2025 ? @ackurth.bsky.social and myself put together a stellar lineup for a satellite workshop on the effects of top-down signals on neuronal dynamics! 🧠✨📣 bernstein-network.de/bernstein-co...

New preprint alert 📣📣📣 New results showing how strong the link between brain dynamics and structure is 💫👀🧪 "Spontaneous spiking statistics form unique area-specific fingerprints and reflect the hierarchy of cerebral cortex" 🐾🧠 www.biorxiv.org/content/10.1...

Spontaneous spiking statistics form unique area-specific fingerprints and reflect the hierarchy of cerebral cortex

The cerebral cortex, from sensory to higher cognitive areas, is hierarchically organised Several dynamical and anatomical measures, such as timescales and neurotransmitter receptor expression, have in...

biorxiv.org

📣 New & better method to compare matrices just dropped! 📣 Ever wanted to compare things like connectivity matrices or large matrices of neuron responses in trials? There is a new way to do so in town ✨🧪 Singular Angle Similarity (SAS) journals.aps.org/prxlife/abst... 🧵

The NEST virtual conference is gearing up! Join us June 17-18, 2025 for a chance to exchange ideas and learn about developments in simulation science! Interested in contributing? The abstract submission deadline has been extended to **April 30**. Visit nest-simulator.org/conference. #nestsim

New preprint: "The geometry of the neural state space of decisions", work by Mauro Monsalve-Mercado, buff.ly/42wVHD5. Surprising results & predictions! (Thread) We analyze neuropixel population recordings in macaque area LIP during a reaction time, random-dot motion 1/