🔥The 9th Round of Easy Loan, Earn $40 Reward is in progress❗️ ⏰ Promotion Period: January 15th - Feburary 15th, 2025 👉 Register now and check more details at gate.io/campaigns/358

Symmetries are everywhere — from butterfly’s wings to Greek temples. But detecting them in noisy data? That’s a challenge. 🦋🏛 Our #SIGGRAPHAsia2024 paper, Robust Symmetry Detection via Riemannian Langevin Dynamics, tackles this: symmetry-langevin.github.io 🧵(1/n)

ISTAustria welcomes renowned scientist, inventor, and entrepreneur Alex Bronstein as a faculty member! After a year as a Visiting Professor, he joins us to advance machine vision, #MachineLearning in life sciences—and even #AI in music! Read more 🔗 bit.ly/48uOcgX

✨ Meet #ResiDual, a novel perspective on the alignment of multimodal latent spaces! Think of it as a spectral "panning for gold" along the residual stream. It improves text-image alignment by simply amplifying task-related directions! 🌌🔍 arxiv.org/abs/2411.00246 [1/6]

Good opportunity to announce Wasserstein Flow Matching (arxiv.org/abs/2411.00698)! New work in collaboration with Brandon Amos & Aram-Alexandre Pooladian on extending Flow Matching to the space of distributions!

Personal Update - I am very happy to start as an Assistant Professor in the Department of Applied Mathematics @ University of Twente! ⭐️ Do reach out for all things: Numerics, Geometry and Machine Learning sites.google.com/view/paigautam…

We attempted to make Normalizing Flows work really well, and we are happy to report our findings in paper arxiv.org/pdf/2412.06329, and code github.com/apple/ml-tarfl…. [1/n]

Next up - Contributed Talk! Generative modeling of protein ensembles guided by crystallographic electron densities Sai Advaith Maddipatla · Nadav Sellam · Sanketh Vedula · Ailie Marx · Alex Bronstein

Generative modeling of protein ensembles guided by crystallographic electron densities 1. A novel density-guided approach to protein ensemble modeling is proposed, directly sampling from the electron density map and recovering intricate multi-modal conformations overlooked by

Inverse problems with experiment-guided AlphaFold. arxiv.org/abs/2502.09372

Very interesting work is also happening using diffusion-based priors! 🔗Solving Inverse Problems in Protein Space Using Diffusion-Based Priors arxiv.org/abs/2406.04239 & Inverse problems with experiment-guided AlphaFold arxiv.org/abs/2502.09372 13/14

Navigating the Latent Space Dynamics of Neural Models. arxiv.org/abs/2505.22785

Representing local protein environments with atomistic foundation models １．This paper introduces a novel method to represent local protein environments using embeddings derived from atomistic foundation models (AFMs). These representations capture both structural and chemical

The first paper citing Egret-1 was released recently, less than a month after we released our model... Sanketh Vedula, Alex Bronstein, + co-workers use Egret-1 embeddings to predict local protein properties like NMR shift + secondary structure.

Excited to share our recent work on NNP representations, done in collaboration with many amazing colleagues!

Today on Rowan's blog: a guest post from Meital Bojan and Sanketh Vedula about how to use embeddings from NNPs to encode the environment about a given atom. These embeddings are very sensitive to local environments and can be used for downstream prediction (NMR, pKa, etc):

Meital and Sanketh describe their recent paper, discuss what the embeddings are learning, and suggest areas for future research. Check out the full post! rowansci.com/blog/local-ato…

Really cool work by Meital and Sanketh Vedula, excited to highlight this. Check out their blog post and paper here!

Our recent work on NNP representations, featured on Rowan's blog!

Our guided AlphaFold work is being presented this week at ICML!