ICML Spotlight 🚨 Equivariance is too slow and expensive, especially when you need global context. It makes us wonder if it even worths the cost, especially in high-dimensional problems? We present Geometric Hyena Networks — a simple equivariant model orders of magnitude more

Clifford Algebra Neural Networks are undeservedly dismissed for being too slow, but they don't have to be! 🚀Introducing **flash-clifford**: a hardware-efficient implementation of Clifford Algebra NNs in Triton, featuring the fastest equivariant primitives that scale.

This work is a continuation of our work on convolutional networks and symmetry breaking. Equivariance helps. So does breaking it — if done right. We explore the advantages of equivariance and controlled symmetry breaking to boost performance across tasks for point clouds.

An interlude about structure computation, SSM and attention. Myosotis: https://t.co/0RXbej9bFm I hope to tell more in this line of work later. See poster at SPIGM workshop.

Postdoc position available in the BioEmu team at @MSFTResearch AI for Science - Berlin DE or Cambridge UK. Looking for candidates with backgrounds in #MachineLearning #AI Biophysics or Bioinformatics https://t.co/DwGuTO1qlf

📣 Our team is hiring (research scientist and SWE roles) - perhaps the best place where you can blend graph learning experience with the most advanced AI research :) Application link and all the downsides of our job: https://t.co/POGyCAnfbZ Drop me a CV in PMs too

📢We are very excited to announce a ✨two-day TAG event ✨(alongside NeurIPS) in San Diego! Join us for an exciting two days filled with incredible speakers, panelists, and multiple poster sessions. 🌟Submission deadline is 1st October AOE. Call for papers is here!

Symmetries+biology is a beautiful combination!

Job alert 🚨 Our team is looking for ML Research Scientist to join Johnson&Johnson research. We work on geometric deep learning and LLMs in drug discovery. 🧬🤓 Drop me a message if you’re interested, or share if you know someone who’s a great fit! Multiple locations available.

Our code for Hierarchical RNA Language models is out! Multiple training regimes, architectures and evaluations, check it out! Code:

A common takeaway from "the bitter lesson" is we don't need to put effort into encoding inductive biases, we just need compute. Nothing could be further from the truth! Better inductive biases mean better scaling exponents, which means exponential improvements with computation.

Cool!

Excited to share Flow Matching Policy Gradients: expressive RL policies trained from rewards using flow matching. It’s an easy, drop-in replacement for Gaussian PPO on control tasks.

🚨 The era of infinite internet data is ending, So we ask: 👉 What’s the right generative modelling objective when data—not compute—is the bottleneck? TL;DR: ▶️Compute-constrained? Train Autoregressive models ▶️Data-constrained? Train Diffusion models Get ready for 🤿 1/n

Great talk

Small poster on large-scale equivariance. Kudos to the presenter @SharvVadgama ⭐️

Happenning today! Come to see our Geometric Hyena poster! 🤓

Excited to be presenting “The Price of Freedom: Exploring Expressivity and Runtime Tradeoffs in Equivariant Tensor Products” at ICML today, at the East Exhibition Hall in Hall A/B at 4:30 PM! Work done with the amazing YuQing Xie, @_MitKotak and @tesssmidt. #ICML2025

Excited to be in Vancouver for #ICML2025 this week! I’m here to talk about our latest work “Low-distortion and GPU-compatible tree embeddings in hyperbolic space”. If you're interested in graph embeddings and hyperbolic geometry, come and check it out! More details below 👇