David Klindt

UniReps Workshop Unifying Representations in Neural Models

Join us for the 3rd edition of the UniReps Workshop at #NeurIPS2025! 🧠🤖

We're bringing together researchers from #AI, #Neuroscience, and #CogSci to explore why different neural models learn similar representations.

📝 Call for Papers is OPEN on OpenReview!
🌐 Website: unireps.org/2025/

📢 We're thrilled to announce that the UniReps workshop will return to @NeurIPSConf 25 for its 3rd edition!

🔵Check our new Call for Papers at: unireps.org/2025/call-fo...

🔴Submit your work at: openreview.net/group?id=Neu...

See you in San Diego!🌴🇺🇸

Simons Collaboration on Ecological Neuroscience The mission of the Simons Collaboration on Ecological NEuroscience (SCENE) is to understand how affordances offered by the world shape representations in the mind and the brain.

Delighted to see our new collaboration on ecological neuroscience 🧩🌿🧠 announced today! A huge ❤️🙏🏼 to @simonsfoundation.org

I’m so excited by this theory-driven question. It’s core to our research on adaptive intelligence & sensorimotor systems 🙌 #SCENE

www.simonsfoundation.org/2025/04/24/s...

I am also surprised, maybe an effort to privatise AI research?

➕ Bonus: Theory can explain the “Platonic Representation Hypothesis”—the striking observation that different models often learn the same representations.

arxiv.org/abs/2405.07987

With the right assumptions (hopefully not too 🧚‍♀️), you get a rigorous mathematical explanation for why this happens 🤓

So—does this mean theory is doomed, and AI engineering is just a random walk? Not at all!

💡 @rpatrik96.bsky.social @wielandbrendel.bsky.social Randall Balestriero did an amazing job clarifying where theory can help practice—and where practice should inspire theory.

🤝

In Theory fairyland:
Infinite data, models converge globally, latents on the right manifold, with the right statistics... 🧚

In Practice:
Noise, Batch size, learning rate, data augmentations, inductive biases, did I mention noise? & all the gritty stuff that actually matters 🛠️

(We still 💙 theory.)

An Empirically Grounded Identifiability Theory Will Accelerate Self-Supervised Learning Research Self-Supervised Learning (SSL) powers many current AI systems. As research interest and investment grow, the SSL design space continues to expand. The Platonic view of SSL, following the Platonic Repr...

🚨 New paper on the limits of identifiability theory!

Mathematical guarantees for neural representations are thrilling—especially in a field driven by trial and error.

But sometimes theory lives in "fairyland" ...

👉 arxiv.org/abs/2504.13101

What about any old system identification paper that shows how predictive performance goes up as a function of data?Not sure it fits the bill of scaling law and foundation model, but it shows pretty much the same thing 🤓

Wow this is really cool, especially since the sensors are so close; and awesome to see the downstream effect on head direction coding

A Perspective from the Ecker lab discusses the progress and challenges of using computer vision approaches for behavior studies of primates in natural environments.

www.nature.com/articles/s41...

Come do cool BCI experiments with wellcome trust fellow Mostafa!

Amazing work! I dreamt about getting this to work at such scale and predicting anatomical cell types from functional data ever since we wrote the what/where readout paper 🤩

The MICrONS Project An unprecedented dataset of high resolution anatomical images of individual cells in mouse visual cortex, mapped on to their responses. This integrated view of function and structure lays a foundation...

After 7 years, thrilled to finally share our #MICrONS functional connectomics results!

We recorded activity from ~75K neurons in the visual cortex of a single mouse, then mapped its wiring using electron microscopy.

nature.com/immersive/d42859-025-00001-w/index.html

Tree Style Tab - Chrome Web Store A tree style tab navigator

have you tried something like chromewebstore.google.com/detail/tree-...

📣 Grad students and postdocs in computational and theoretical neuroscience: please consider applying for the 2025 Flatiron Institute Junior Theoretical Neuroscience Workshop! All expenses are covered. Apply by April 14. jtnworkshop2025.flatironinstitute.org

👀

We're bringing together leading researchers from top institutions worldwide to Split, Croatia this summer. If you're passionate about brain-inspired AI, computational neuroscience, or fundamental mathematical principles driving intelligence, this conference is the place to be!

Going to #Cosyne2025?

Join us (w/ @mdiamantaki.bsky.social) for our exciting workshop

Object-centric Neuronal Representations

toliaslab.org/workshop/cos...

Featuring diverse species💁‍♀️🐒🐁🐝🕷️🤖

Our goal: Discuss universal & species-specific principles of object-centric vision

See you in Montreal🥳

Goodhart's law; Kept on pushing, unsuccessfully, to have that mentioned in here proceedings.neurips.cc/paper_files/...

Party poster for dance party on final night of Cosyne 2025 workshops. It will take place April 1st, 2025, 10PM to 3AM at Le P'tit Caribou.

Coming to the #Cosyne2025 workshops? Wanna dance on the final night? We got you covered.

@glajoie.bsky.social and I have organized a party in Tremblant. Come and get on the dance floor y'all. 🕺

April 1st
10PM-3AM
Location: Le P'tit Caribou

DJs Mat Moebius, Xanarelle, and Prosocial

Please share!

Visa/ green card holders in the US who had planned to go to @cosynemeeting.bsky.social , what are your current thoughts about leaving and reentering the US / has your university issued any new directive? I'm happy to dm/ chat over Signal too. #cosyne25 #neuroskyence

Mathematical Methods in Computational Neuroscience Summer school in Eresfjord, Norway (July 8th - 26th, 2024)

Applications are open for the summer school 𝗠𝗮𝘁𝗵𝗲𝗺𝗮𝘁𝗶𝗰𝗮𝗹 𝗠𝗲𝘁𝗵𝗼𝗱𝘀 𝗶𝗻 𝗖𝗼𝗺𝗽𝘂𝘁𝗮𝘁𝗶𝗼𝗻𝗮𝗹 𝗡𝗲𝘂𝗿𝗼𝘀𝗰𝗶𝗲𝗻𝗰𝗲.
compneuronrsn.org

Application deadline: March 31, 2025
Located in beautiful Eresfjord, Norway 🇳🇴

#Neuroscience #ComputationalNeuroscience #KavliNeuro
@kavlintnu.bsky.social

Kudos to the great team @ninamiolane.bsky.social @rpatrik96.bsky.social @charlesoneill.bsky.social Harald Maurer 🙏

Honestly, the real achievement? We managed to sneak pics of both our pets into the paper. 😝 🐾
Check it out and let us know what you think!
arxiv.org/abs/2503.01824

From superposition to sparse codes: interpretable representations in neural networks Understanding how information is represented in neural networks is a fundamental challenge in both neuroscience and artificial intelligence. Despite their nonlinear architectures, recent evidence sugg...

We’re looking for feedback and especially criticism before sending this off to a journal. If you’re into neural representations, we’d love to hear your thoughts! 📝🔥
arxiv.org/abs/2503.01824

At the core:
1️⃣ Identifiability theory
2️⃣ Compressed sensing
3️⃣ Quantitative interpretability

Our goal is a unified model for LRH, superposition, sparse coding, and AutoInterp—backed by theory and practical insights. 🧠🔍
arxiv.org/abs/2503.01824

This all started last year with a simple question: who first applied sparse coding to neural representations for more interpretable codes? That question led us to uncover links between identifiability, compressed sensing, and interpretability—a story that was too good not to tell. 🧩

From superposition to sparse codes: interpretable representations in neural networks Understanding how information is represented in neural networks is a fundamental challenge in both neuroscience and artificial intelligence. Despite their nonlinear architectures, recent evidence sugg...

🔵 New paper! We explore sparse coding, superposition, and the Linear Representation Hypothesis (LRH) through identifiability theory, compressed sensing, and interpretability. If you’re curious about lifting neural reps out of superposition, this might interest you! 🤓
arxiv.org/abs/2503.01824