Check out our new paper on computational modelling of epigenetic plasticity! With the lab of @graeffjohannes.bsky.social academic.oup.com/brain/advanc...

A computational framework for epigenetic plasticity in memory

Delamare et al. present a computational model of memory formation inspired by recent experimental data, in which individual memories are encoded not only i

academic.oup.com

Update on this work! We've extended our transitive inference study to large language models 🧵

Jesse Geerts@jessegeerts.bsky.social · last yr.

🧠 How do transformers learn relational reasoning? We trained small transformers on transitive inference (if A>B and B>C, then A>C) and discovered striking differences between learning paradigms. Our latest work reveals when and why AI systems generalize beyond training data 🤖

I’m excited to share my first PhD preprint!🎉 We studied how interactions between medial entorhinal cortex (MEC) and hippocampus shape theta sequences during navigation, and asked whether some “planning-like” patterns in hippocampus could arise from upstream MEC dynamics. (1/8)

Check out our new work on motor learning across multiple brain regions!

Jesse Geerts@jessegeerts.bsky.social · 7mo ago

🧠 New year, new preprint! Why does motor learning involve multiple brain regions? We propose that the cortico-cerebellar system learns a "map" of actions where similar movements are nearby, while basal ganglia do RL in this simplified space. www.biorxiv.org/content/10.6...

We were asked to write a commentary on NeuroAI for Nat. Rev. Neuro. with Sadra Sadeh, so here it is: www.nature.com/articles/s41...

Trying to train RNNs in a biol plausible (local) way? Well, try our new method using predictive alignment. Paper just out in Nat. Com. Toshitake Asabuki deserves all the credit! www.nature.com/articles/s41...

Are predictions computed hierarchically, or can they be computed locally? Check out our paper! Congrats, Toshitake Asabuki and @colleenjg.bsky.social ! www.pnas.org/doi/10.1073/...

Check out our new paper how to embed state transitions in recurrent networks. Well done Toshitake Asabuki! elifesciences.org/articles/95243

Embedding stochastic dynamics of the environment in spontaneous activity by prediction-based plasticity

A computational model shows how synaptic plasticity shapes spontaneous activity to encode the transition statistics of sensory experience.

elifesciences.org

Check out our new work on in context learning! With @jessegeerts.bsky.social @neurokim.bsky.social @scychan.bsky.social

Jesse Geerts@jessegeerts.bsky.social · last yr.

🧠 How do transformers learn relational reasoning? We trained small transformers on transitive inference (if A>B and B>C, then A>C) and discovered striking differences between learning paradigms. Our latest work reveals when and why AI systems generalize beyond training data 🤖


We're launching the data collection phase of the Community Predictive Processing project now that our plans have been published on arXiv 👉 arxiv.org/abs/2504.09614

Neural mechanisms of predictive processing: a collaborative community experiment through the OpenScope program

This review synthesizes advances in predictive processing within the sensory cortex. Predictive processing theorizes that the brain continuously predicts sensory inputs, refining neuronal responses…

arxiv.org

How should you combine sensory inputs with your internal predictions? Well, use uncertainty directly computed by your prediction-error circuits. Cool? Check out our Nat. Com. paper with Loreen Hertäg and @k47h4.bsky.social . www.nature.com/articles/s41...

Uncertainty estimation with prediction-error circuits - Nature Communications

How the brain integrates sensory input and predictions to adapt to change is not fully understood. Here authors build a neural network model to show how prediction-error neurons compute uncertainty of...

nature.com

The Clopath Lab has a fully funded PhD position open for Oct in Computational Neuroscience at Imperial College London. If you are interested, just send us an informal e-mail!

Check out our new method to optimize neural representations from behaviour!

Sainsbury Wellcome Centre@sainsburywellcome.bsky.social · last yr.

“SIMPL: Scalable and hassle-free optimisation of neural representations from behaviour” openreview.net/forum?id=9kF... From @tomnotgeorge.bsky.social Pierre Glaser, Kim Stachenfeld, @caswell.bsky.social @clopathlab.bsky.social

Want to hear more about how feedback can guide learning in RNNs for motor adaptation. Here is our new paper in Nat. Com. with Barbara Feulner and @juangallego.bsky.social www.nature.com/articles/s41...

A neural implementation model of feedback-based motor learning - Nature Communications

How the brain adapts our movements to new conditions remains unclear. Here, the authors show that a recurrent neural network that controls its output using error-based feedback can learn to count...

nature.com

Check out our new paper with the Margrie lab in Cell on motor and vestibular signals in the visual cortex. Well done Mateo Velez-Fort!

Sainsbury Wellcome Centre@sainsburywellcome.bsky.social · last yr.

Check out the full paper in @cp-cell.bsky.social from Mateo Velez-Fort, Lee Cossell, Laura Porta, @clopathlab.bsky.social and Troy Margrie www.cell.com/cell/fulltex...

Want to hear more about hippocampus-to-cortex feedback circuit? Our paper with the Basu lab is out in Nat Neuro! rdcu.be/eak1k . Well done Tanvi Butola!

Hippocampus shapes entorhinal cortical output through a direct feedback circuit

Nature Neuroscience - Distinct hippocampal feedback pathways to deep and superficial layers of the cortex in mice differentially modulate excitation–inhibition dynamics, plasticity and...

rdcu.be

Read the full paper in @science.org from Sara Mederos, Patty Blakely, @nicolevissers1.bsky.social @clopathlab.bsky.social and @sonjahofer.bsky.social ⬇️ www.science.org/doi/10.1126/...

Overwriting an instinct: Visual cortex instructs learning to suppress fear responses

Fast instinctive responses to environmental stimuli can be crucial for survival but are not always optimal. Animals can adapt their behavior and suppress instinctive reactions, but the neural pathways...

science.org

See our new paper in Science with the Hofer lab. Well done Sara Mederos!

Sonja Hofer@sonjahofer.bsky.social · last yr.

So excited to see this out, heroic effort by the brilliant Sara Mederos, uncovering the detailed brain mechanisms of how animals learn to overcome their instinctive fear responses, from neural circuits & cell types down to synaptic plasticity mechanism! www.science.org/doi/10.1126/...

So excited to see this out, heroic effort by the brilliant Sara Mederos, uncovering the detailed brain mechanisms of how animals learn to overcome their instinctive fear responses, from neural circuits & cell types down to synaptic plasticity mechanism! www.science.org/doi/10.1126/...

Overwriting an instinct: Visual cortex instructs learning to suppress fear responses

Fast instinctive responses to environmental stimuli can be crucial for survival but are not always optimal. Animals can adapt their behavior and suppress instinctive reactions, but the neural pathways...

science.org

Happy to see three computational neuroscience awards among the ERC Consolidator grants announced today, and even more impressively 3/3 are women: @clopathlab.bsky.social, @gjorjulijana.bsky.social and Alex Cayco-Gajic All top researchers, delighted for them. Congrats erc.europa.eu/sites/defaul...

Very grateful for ERC to fund my research for the next 5 years! erc.europa.eu/news-events/...

Consolidator Grants: ERC awards €678m in grants to back excellent research across Europe

The European Research Council (ERC) has awarded its 2024 Consolidator Grants to 328 researchers across Europe. These grants, totalling €678 million, aim to support outstanding scientists and scholars ...

erc.europa.eu

What are the brain’s “real” tuning curves? Our new preprint "SIMPL: Scalable and hassle-free optimisation of neural representations from behaviour” argues that existing techniques for latent variable discovery are lacking. We suggest a much simpl-er way to do things. 1/21🧵

This paper from Claudia Clopath's lab is very exciting for me. I've been musing for a while now that the challenge of approximating gradient descent is probably made way easier in the brain by feedback signals for control. Here they show it!!! openreview.net/forum?id=xav... 🧠📈 #NeuroAI 🧪

Feedback control guides credit assignment in recurrent neural networks

How do brain circuits learn to generate behaviour? While significant strides have been made in understanding learning in artificial neural networks, applying this knowledge to biological networks...

openreview.net