📢 Join us for the next UniReps x @ellis.eu speaker series even happening on 29th April at 4:00 PM CET with João Sacramento and Paul Friedrich!👾

Keynote Speaker Nomination | ELLISxUniReps Speaker Series Hi all! We are collecting nominations for keynote speakers for the ELLISxUniReps Speaker Series! If you know a senior researcher whose work aligns with our community’s focus on representation learnin...

🔵🔴You want to see a specific presenter at the ELLIS x UniReps Speaker Series? We hear you! 👀

Fill in this form to nominate keynote speaker candidates:
🔗 forms.gle/rQjkZ1YVWLwj...

We’re starting in 15 minutes! 🎉

Join us tomorrow on zoom or in person at ETH Zurich!

When, how, and why do neural models learn similar representations?

📢 The next ELLIS x @unireps.bsky.social speaker series event is with Antonio Orvieto and Sara Kangaslahti. Don’t miss it!

📍 Virtual
📅 Feb 26
🕓 16:00 CET
🔗 https://bit.ly/4aED6rr

📢The next UniReps x @ellis.eu speaker series event is happening on 26 February at 4:00 PM CET with Antonio Orvieto and Sara Kangaslahti. Don’t miss it!🚀

That was fun! Thanks for the invitation :-)

Join us in one hour! 📣 @noagarciad.bsky.social @thwiedemer.bsky.social

Zoom link: ethz.zoom.us/j/66426188160

📢The first UniReps x @ellis.eu speaker series in 2026 is coming!

Come join us on 15th January at 1 pm CET with
@noagarciad.bsky.social and @thwiedemer.bsky.social !

Starting in 30 minutes! Join us on zoom!

Join us tmr!

📢The UniReps x @ellis.eu
speaker series is back! Come join us in our next appointment 18th December 4 pm CET with @meenakshikhosla.bsky.social
and Raj Magesh Gauthaman🔵🔴

🔵🔴 We are now at the last invited talk of today! Eva Dyer from the University of Pennsylvania is presenting “Learning Who’s Who in the Brain: Population Context as a Key to Neuronal Identity”

Last talk of the session: “Self-Supervised Learning from Structural Invariance” by Jungyoon Lee

Fourth talk: “Sparse Autoencoder Neural Operators: Model Recovery in Function Spaces” by Bahareh Tolooshams and Ailsa Shen

Third talk: "Weight Weaving: Parameter Pooling for Data-Free Model Merging" by Levy Chaves

Second talk: Shared Parameter Subspaces and Cross-Task Linearity in Emergently Misaligned Behavior by Daniel Aarao Reis Arturi ✨

The second session of contributed talks is kicking off now.
First talk coming up: SemCLIP: A Semantic Memory Aligned Vision Language Model by Tanveer Syeda-Mahmood 🚀

We now continue after the break with Randall Balestriero from Brown University, presenting “LeJEPA: Provable and Scalable Self-Supervised Learning Without the Heuristics ” 👏

Up next, our Mentorship Panel featuring
@abeirami.bsky.social, @sarahooker.bsky.social, and @meenakshikhosla.bsky.social moderated by @icannistraci.bsky.social.

Excited to have such inspiring leaders sharing their perspectives on the current state of the ML peer-review system! 🔍✨

Fifth and final talk of Session 1: "On Defining Neural Averaging" by Su Hyeong Lee 🧠⚡️

Fourth talk: "Superposition disentanglement of neural representations reveals hidden alignment" by André Longon 💡🌟

Third talk: "Universally Converging Representations of Matter Across Scientific Foundation Models" by Sathya Edamadaka and Soojung Yang 📊⚡️

Second talk: "Neuron-Level Linguistic Selectivity in LLMs via Minimal Pairs: A Neuroscience-Inspired Framework" by Linyang He 🧠✨

Session 1 of the Contributed Talks start now! 🚀

First talk: Neural Embedding Alignment Reveals Nonlinear Latent Transformations across Brain Regions by Stephen Keeley

Did you know that you could create hybrids across different neural net architectures—*without* updating the weights??

Come check out my poster at the UniReps @unireps.bsky.social workshop at 3:45pm today at #NeurIPS! (Or you can already browse the posters on the wall in 20D throughout the day.) 😃

I thought this was a fascinating way to motivate model merging: to be able to *interpolate* between model classes, rather than be forced into discrete choices. (models trained on different datasets, models of different sizes, etc.)

"Interpolation and alignment are two sides of the same coin."

It’s our pleasure to welcome @dmelis.bsky.social from Harvard SEAS to UniReps 2025, presenting “From Discrete to Continuous: Interpolating Models, Datasets, and Representations!” 🔁📊

Join us in Ballroom 20D for more exciting discussions on representational alignment!

We now continue with Prof. Nikolaus Kriegeskorte from Columbia University, presenting "Meaningful and Powerful Model Comparisons"!