Under He/Lecun inits, theory implies Kernel OR Unstable regimes as width→∞. Discrepancies (e.g. feature learning) are seen as finite width effects. Our #NeurIPS2025 spotlight refutes this: practical nets do not converge to kernel limits; Feature learning persists as width→∞🧵

At #NeurIPS ? Visit our posters! 🧵 Demystifying Spectral Feature Learning for Instrumental Variable Regression: #2600, Wed 11am Regularized least squares learning with heavy-tailed noise is minimax optimal: #3012, Wed 4:30pm ✨spotlight✨ 1/2

Find us at NeurIPS, Thur 4:30 pm #2115! We know networks have to be both plastic and stable but we're used to thinking about computations, such as memory, as additional requirements. Instead, we find that almost all stable & plastic networks display simple memory abilities.

👋 Will you be at #NeurIPS? Stop by our researchers' posters to learn more about their work!

The date should of course be August 17-28 2026!

📢 Project Research Internship (spring 2026) Apply now to work with Dr Leena Chennuru Vankadara (@leenacvankadara.bsky.social) on the fundamental principles governing the behaviour of modern large-scale machine learning systems. ⏰ Deadline: 18 Dec 2025, 1pm GMT ℹ️ www.ucl.ac.uk/life-science...

✈️ Heading to #SfN25? Check out Kira's (@kiradust.bsky.social) work "Autistic individuals learn to exploit temporal regularities in simple but not complex contexts" at Early Career Poster Session (Nov 15 18:45-20:45, A21) or Autism: Behavioral Analysis session (Nov 16 13-17, A27). Don't miss it!

If you are interested in Bonsai and attending #SfN25 in San Diego, come and hang out by the Bonsai Foundation Booth 3726, Hall B2. Task logic, Harp devices, online real-time Machine Learning, new language features. We will be there everyday with live demos and open to chat about all things Bonsai!

Honored to have a research highlight featuring our work! A comprehensive overview of our results and their impact for future research and applications: www.nature.com/articles/s41...

💬 AI for Good Webinar Series 'From Molecules to Models' - today with @arthurgretton.bsky.social (@ucl.ac.uk, ELLIS Fellow) on causal effect estimation. 📅 Oct 29th 🕓 17:00-18:00 CET 🔗 Registration: aiforgood.itu.int/event/learn-... ℹ️ More info: ellis.eu/news/continu...

⏰ Deadline approaching for our PhD programmes! 👋 Apply by 3 Nov to join us and immerse yourself in world-class research and vibrant London life! Gatsby Unit: ucl.ac.uk/life-science... @sainsburywellcome.bsky.social: sainsburywellcome.org/web/content/... Joint: www.ucl.ac.uk/life-science...

Looking forward to next week's Winter School on Causality and Explainable AI! xai-winter-school.github.io

Applying to the #ELLISPhD Program? If you are selecting a Gatsby Unit faculty member as your main supervisor, don't forget to apply to the Gatsby Unit PhD Programme in parallel. Our deadline is 3 Nov 2025. 👉 www.ucl.ac.uk/life-science...

Congratulations to all the new SWC & @gatsbyucl.bsky.social PhD students for successfully completing Boot Camp! We are delighted to welcome you to the community. Students showcased their robots in our hotly anticipated annual @noblackboxes.bsky.social Robot Competition 🦾⚽💻

🙋Are you interested in bridging theory & experiments? Applications are now open for 2026 entry to the Gatsby Unit & SWC joint PhD programme. Join us and be part of a vibrant research community! 💰 Fully-funded 4-year programme ℹ️ www.ucl.ac.uk/life-science... @sainsburywellcome.bsky.social

Interested in a #PhD in machine learning & theoretical neuroscience? Applications for 2026 entry to the Gatsby Unit PhD programme are open! 💰 Fully-funded 4-year programme 🤝 Close links to @sainsburywellcome.bsky.social ℹ️ www.ucl.ac.uk/life-sciences/gatsby/study-and-work/gatsby-unit-phd-programme

A complete brain-wide activity map at single-cell resolution has been revealed for the first time. Researchers recorded from 650,000+ neurons across 279 brain areas to track decision-making in mice. Read the story: 🔗 www.sainsburywellcome.org/web/research...

Made possible only by the unique environment and amazing community at Janelia and the Gatsby. @hhmijanelia.bsky.social @gatsbyucl.bsky.social Feedback, ideas, or requests for transgenic lines, protocols, and code are warmly welcome!

ProbNum 2025 Keynote 2 ``Gradient Flows on the Maximum Mean Discrepancy'' by @arthurgretton.bsky.social ( @gatsbyucl.bsky.social and Google DeepMind. Slides available here: probnum25.github.io/keynotes

📢 ❗JOB OPPORTUNITY: Today is the closing date for this fantastic role we are funding at @gatsbyucl.bsky.social. ⏰ Apply before 23.59BST. 👇

Aeon is a new platform for continuous brain and behaviour research in naturalistic environments — already used in 15+ projects, including long-term social learning studies. 🔗 www.sainsburywellcome.org/web/blog/new... 📄 Preprint: www.biorxiv.org/content/10.1...

Are similar representations in neural nets evidence of shared computation? In new theory work w/ Lukas Braun (lukasbraun.com) & @saxelab.bsky.social, we prove that representational comparisons are ill-posed in general, unless networks are efficient. @icmlconf.bsky.social @cogcompneuro.bsky.social

Main takeaway: Valid representational comparison relies on implicit assumptions (task-optimization *plus* efficient implementation). ⚠️ More work to do on making these assumptions explicit! 🧠 CCN poster (today): 2025.ccneuro.org/poster/?id=w... 📄 ICML paper (July): icml.cc/virtual/2025/poster/44890

Our #CCN2025 GAC debate w/ @gretatuckute.bsky.social, Gemma Roig (www.cvai.cs.uni-frankfurt.de), Jacqueline Gottlieb (gottlieblab.com), Klaus Oberauer, @mschrimpf.bsky.social &‬ @brittawestner.bsky.social asks: 📊 What benchmarks are useful for cognitive science? 💭 2025.ccneuro.org/gac

Sequential kernel embedding for mediated and time-varying dose response curves Appearing in Bernoulli: projecteuclid.org/journals/ber... with preprint here: arxiv.org/abs/2111.03950 ...along with code! github.com/liyuan9988/K... Rahul Singh, Liyuan Xu

JOB: We're recruiting for a 2-year post doc fellowship based at @gatsbyucl.bsky.social working with Prof Arthur Gretton to create new statistically efficient causal and interaction models with the aim of elucidating cellular interactions ⏰closing 27 August 👉 details: www.ucl.ac.uk/work-at-ucl/...

Does the brain learn by gradient descent? It's a pleasure to share our paper at @cp-cell.bsky.social, showing how mice learning over long timescales display key hallmarks of gradient descent (GD). The culmination of my PhD supervised by @laklab.bsky.social, @saxelab.bsky.social and Rafal Bogacz!

New from @sainsburywellcome.bsky.social, @gatsbyucl.bsky.social, NeuroGEARS & DataJoint: Aeon - an open-source platform combining long-term neural recording with naturalistic behavioural tracking in enriched habitats. 📑 Preprint: www.biorxiv.org/content/10.1...

🎓Thrilled to share I’ve officially defended my PhD!🥳 At @gatsbyucl.bsky.social, my research explored how prior knowledge shapes neural representations. I’m deeply grateful to my mentors, @saxelab.bsky.social and @caswell.bsky.social, my incredible collaborators, and everyone who supported me!

🥳 Congratulations to Rodrigo Carrasco-Davison on passing his PhD viva with minor corrections! 🎉 📜 Principles of Optimal Learning Control in Biological and Artificial Agents.

Come chat about this at the poster @icmlconf.bsky.social, 11:00-13:30 on Wednesday in the West Exhibition Hall #W-902!

Excited to share new work @icmlconf.bsky.social by Loek van Rossem exploring the development of computational algorithms in recurrent neural networks. Hear it live tomorrow, Oral 1D, Tues 15 Jul West Exhibition Hall C: icml.cc/virtual/2025... Paper: openreview.net/forum?id=3go... (1/11)

Distributional diffusion models with scoring rules at #icml25 Fewer, larger denoising steps using distributional losses! Wednesday 11am poster E-1910 arxiv.org/pdf/2502.02483 @vdebortoli.bsky.social Galashov Guntupalli Zhou @sirbayes.bsky.social @arnauddoucet.bsky.social

Accelerated Diffusion Models via Speculative Sampling, at #icml25 ! 16:30 Tuesday July 15 poster E-3012 arxiv.org/abs/2501.05370 @vdebortoli.bsky.social Galashov @arnauddoucet.bsky.social

👋 Attending #ICML2025 next week? Don't forget to check out work involving our researchers!

🎊 Congratulations to Kevin Huang (@kevhhuang.bsky.social‬) on passing his PhD viva with minor corrections! 🥳 🔖 Universality beyond the classical asymptotic regime

It was a pleasure to welcome work experience students to SWC and @gatsbyucl.bsky.social this week, through the Camden STEAM initiative with @ucllifesciences.bsky.social They explored our labs and learned about the many paths into #neuroscience www.sainsburywellcome.org/web/content/...

How do task dynamics impact learning in networks with internal dynamics? Excited to share our ICML Oral paper on learning dynamics in linear RNNs! with @clementinedomine.bsky.social @mpshanahan.bsky.social and Pedro Mediano openreview.net/forum?id=KGO...

🥳 Congratulations to Clementine Domine (@clementinedomine.bsky.social) on passing her PhD viva with minor corrections! 🎉 🔖 Balancing Learning Regimes: The Impact of Prior Knowledge on the Dynamics of Neural Representations

Dimitri Meunier, Antoine Moulin, Jakub Wornbard, Vladimir R. Kostic, Arthur Gretton Demystifying Spectral Feature Learning for Instrumental Variable Regression https://arxiv.org/abs/2506.10899

New research shows long-term learning is shaped by dopamine signals that act as partial reward prediction errors. The study in mice reveals how early behavioural biases predict individual learning trajectories. Find out more ⬇️ www.sainsburywellcome.org/web/blog/lon...

Our #Artist-in-Residence, Mary Yacoob, has produced new large-scale #cyanotype prints reimagining of #scientific visual data from @ucllifesciences.bsky.social #researchers. Faculty private #exhibition details here: tinyurl.com/YacoobUCL #artandscience #art #artwork @ucl.ac.uk @ucl-cdb.bsky.social

Last but not least of my travel updates: Courtesy of the very kind @slchau.bsky.social, I'm giving a talk at 14:30, 27 Jun @NTU CCDS in SG on data augmentation & Gaussian universality, which strings together several works over my PhD. If you're in SG/Lyon over the next few weeks, let me know! :)

Meanwhile, excited to be in #Lyon for #COLT2025, with a co-first author paper (arxiv.org/abs/2502.15752) with the amazing team -- Matthew M Mallory and our advisor Morgane Austern! Keywords: Gaussian universality, dependent data, convex Gaussian min-max theorem, data augmentation!

Sadly missing ICML due to visa :'(, but looking forward to share our ICML paper (arxiv.org/abs/2502.05318) as a poster at #BayesComp, Singapore! Work on symmetrising neural nets for schrodinger equation in crystals, with the amazing Zhan Ni, Elif Ertekin, Peter Orbanz and Ryan P. Adams

Composite Goodness-of-fit Tests with Kernels, now out in JMLR! www.jmlr.org/papers/v26/2... Test if your distribution comes from ✨any✨ member of a parametric family. Comes in MMD and KSD flavours, and with code. @oscarkey.bsky.social @fxbriol.bsky.social Tamara Fernandez

We’ve successfully automated part of the (neuro)scientific process. Now I may be out of a job. After a lot of iterations, here’s our framework for automatic discovery, built to embrace degeneracy and the realities of underconstrained modeling.

How does in-context learning emerge in attention models during gradient descent training? Sharing our new Spotlight paper @icmlconf.bsky.social: Training Dynamics of In-Context Learning in Linear Attention arxiv.org/abs/2501.16265 Led by Yedi Zhang with @aaditya6284.bsky.social and Peter Latham

Dimitri Meunier, Zikai Shen, Mattes Mollenhauer, Arthur Gretton, Zhu Li Optimal Rates for Vector-Valued Spectral Regularization Learning Algorithms https://arxiv.org/abs/2405.14778

🚨 New paper accepted at SIMODS! 🚨 “Nonlinear Meta-learning Can Guarantee Faster Rates” arxiv.org/abs/2307.10870 When does meta learning work? Spoiler: generalise to new tasks by overfitting on your training tasks! Here is why: 🧵👇