We have a new distillation method that actually *improves* upon its teacher. Moment Matching distillation (arxiv.org/abs/2406.04103) creates fast stochastic samplers by matching data expectations between teacher and student. Work with Emiel Hoogeboom Jonathan Heek @tejmensin. 1/4

Learning to sample: time-reversibilty meets Metropolis-Hastings. Check our #icml2024 paper where we propose an objective that upper-bounds the TV distance between the stationary distribution of a parametric Markov kernel, and the distribution you want to sample from! (1/6)

Coming to devour your 3D anatomies: 🥼LaB-GATr🐊, geometric algebra transformers for large, biomedical surface and volume meshes. Will be presented MICCAI and Gram Workshop. Joint work with Baris Imre, Pim de Haan and Jelmer Wolterink. (code) github.com/sukjulian/lab-…

📢 In our new UAI 2025 paper, we do neural optimal transport with costs defined by a Lagrangian (e.g., for physical knowledge, constraints, and geodesics) Paper: arxiv.org/abs/2406.00288 JAX Code: github.com/facebookresear… (w/ A. Pooladian, C. Domingo-Enrich, Ricky T. Q. Chen)

I will be at ICML Conference next week presenting 2 papers and 2 workshop papers. I'm also on the job market! Please reach out if you're interested in discussing generative models, geometric DL, or AI4Science.

Looking forward to a full week with AABI and ICML Conference starting this Sunday! 🌟I am co-organising AABI on July 21 approximateinference.org ✨Presenting a paper (Neural Diffusion Models), and several workshop contributions during #ICML2024. Reach out if you want to chat!

Very cool to see Aurora mentioned in this Dutch news article on AI weather forecasting! 🔥

Hello #GRaM enthusiasts, Sadly our previous twitter account has been compromised, and we won't be able to get hold of it soon. Hence, we have moved to our new handle GRaM Workshop at ICML 2024. Kindly unfollow Gram Workshop and follow us GRaM Workshop at ICML 2024 for more information. ICML Conference

🐊 LaB-GATr is a transformer neural network designed for large-scale biomedical surface. It is inspired from GATr (pronounced gator 🐊), the Geometric Algebra Transformer (GATr) that cleverly represents inputs and states using projective geometric algebra. A much needed tutorial!

Come to see our poster on GRaM Workshop at ICML 2024 Schubert 1-3 today at 16-17:00 🚀

Looks at this stars 💫⭐️🌟! Anna Kuzina and Babak Esmaeili at want workshop sharing how to construct optimizer if you are Bayesian guy!

Thrilled to receive the outstanding paper award together with Mangal Prakash for our work on SE(3)-Hyena arxiv.org/abs/2407.01049 !🤘 Go long-convolutions!

Had a great time presenting our latest paper on hybrid modeling of broad-front bird migration at the AI for Science workshop at #ICML24 We can now make detailed and interpretable forecasts of departure, flight, and landing at the continental scale! 📰➡️ openreview.net/pdf?id=oAmxqO1…

Exciting News from Chatbot Arena! Google DeepMind's new Gemini 1.5 Pro (Experimental 0801) has been tested in Arena for the past week, gathering over 12K community votes. For the first time, Google Gemini has claimed the #1 spot, surpassing GPT-4o/Claude-3.5 with an impressive

📰 blogpost: maxxxzdn.github.io/blog/cscnns.ht… 🕹️ google colab: colab.research.google.com/drive/1M196l6X… I tried to make the blog post more accessible than the paper and added a lot of supporting visualizations. Please check it out if you are curious about spacetime-equivariant CNNs 🚀

"There's a STAResNet waiting in the sky..." ⭐STAResNet⭐, or "on the importance of choosing the right algebra in #GA networks", is out now on ArXiv and live tomorrow at #AGACSE 2024 in Amsterdam! paper: t.ly/jEJpy code: t.ly/6rMNx more below:

I’m hiring a postdoc to work with me on exciting projects in generative modelling (AI) and/or uncertainty quantification. You'll be part of a great team, embedded in UvA AMLab and the UvA-Bosch Delta Lab. Apply here: vacatures.uva.nl/UvA/job/Postdo… RT appreciated! #ML #GenAI

I expressed more-or-less the same opinion in my first lecture yesterday :) _Don't_ think of GDL as a mechanism to exploit symmetries of data which has a "spatial" geometry. Think of it as a language to express neural network constraints!

Really awesome to see the full potential of real-time generative modeling being realized! Helping to enable this has been a driving goal of our research in efficient generative models for a long time!