I’m stoked to share our new paper: “Harnessing the Universal Geometry of Embeddings” with jack morris, Collin Zhang, and Vitaly Shmatikov. We present the first method to translate text embeddings across different spaces without any paired data or encoders. Here's why we're excited: 🧵👇🏾

Can time series #FoundationModels like Chronos zero-shot generalize to unseen #DynamicalSystems (DS)? No, they cannot. But *DynaMix* can, the first FM based on principles of DS reconstruction, capturing the long-term evolution of out-of-domain DS: arxiv.org/pdf/2505.13192… (1/6)

Nature The pivot penalty in research nature.com/articles/s4158…

Demanding that "everyone should immediately evacuate" Tehran is astonishing. Most of the city’s 10 million residents simply cannot leave. No one can justify striking them by saying: "We warned you!" Warnings mean nothing when escape isn’t an option.

It won't be long and physics will be overrun by AI slop, mark my words. It's a disaster waiting to happen for academia that has been optimizing for quantity over quality for decades, you can basically see it coming. x.com/getjonwithit/s…

AI at least currently, is simply a prediction machine, it is hardly a bad thing to be able to predict the next words/bits. We would be less negative if we differentiate between science and scientists, very likely we will be doing better science, just need to retrain ourselves

Clarifying the conceptual dimensions of representation in neuroscience. arxiv.org/abs/2403.14046

A statistical physics framework for optimal learning Authors: Francesca Mignacco, Francesco Mori arxiv.org/abs/2507.07907

Compression is the heart of intelligence From Occam to Kolmogorov—shorter programs=smarter representations Meet KARL: Kolmogorov-Approximating Representation Learning. Given an image, token budget T & target quality 𝜖 —KARL finds the smallest t≤T to reconstruct it within 𝜖🧵

Excited to co-organize our NeurIPS 2025 workshop on Foundation Models for the Brain and Body! We welcome work across ML, neuroscience, and biosignals — from new approaches to large-scale models. Submit your paper or demo! 🧠💪

prediction ≠ understanding foundation model ≠ theory

Got prov. approval for 2 major grants in Neuro-AI & Dynamical Systems Recons., on learning & inference in non-stationary environments, OOD generalization, and DS foundation models. To all AI/math enthusiasts: Expect job announcements (PhD/PostDoc) soon! Feel free to get in touch.

📢 Present your NeurIPS paper in Europe! Join EurIPS 2025 + ELLIS UnConference in Copenhagen for in-person talks, posters, workshops and more. Registration opens soon; save the date: 📅 Dec 2–7, 2025 📍 Copenhagen 🇩🇰 🔗eurips.cc #EurIPS EurIPS Conference

🌀 Universality in Computable Dynamical Systems: Old and New Can dynamical systems compute? From Turing-complete fluids to Topological Kleene Field Theories, we explore how dynamics encodes computation—past, present, and future. arxiv.org/abs/2507.10725

The modern university is a place where a scientist asking hard questions is seen as a threat, and an administrator burying the truth is called a leader.

Our new preprint compares naïve baselines, network models (incl. PLRNN-based SSMs), and Transformers on 3x40‑day EMA+EMI datasets. PLRNNs gave the most accurate forecasts, yielded interpretable networks, and flagged “sad” & “down” as top leverage points. doi.org/10.1101/2025.0…

Can AI help understand how the brain learns to see the world? Our latest study, led by Joséphine Raugel from FAIR at AI at Meta and École normale supérieure | PSL, is now out! 📄 arxiv.org/pdf/2508.18226 🧵 A thread:

Sampling is hard b/c target distribution can be high-dim with many modes. ML can help, even when state space is discrete (thus non-differentiable)! arxiv.org/abs/2508.10684 constructs a strong sampler by fine tuning a discrete diffusion model via stochastic optimal control / RL.

Our #DynamicalSystems #FoundationModel was accepted to #NeurIPS2025 with outstanding reviews (6555) – first model which can *0-shot*, w/o any fine-tuning, forecast the *long-term statistics* of time series provided a context. Test it on #HuggingFace: huggingface.co/spaces/Durstew… ...

I encourage you to read this article, in which we describe the current situation and the directions in which, in our view, mathematics is heading. Many thanks to Ken Ono for including me in this extraordinary project. I look forward to a wide-ranging discussion and will be