Want to make your model equivariant to certain known transformations without changing the network architecture? Please attend our contributed talk at 11AM Symmetry and Geometry in Neural Representations to know how we achieve this by adding a simple shallow network that learns to canonicalize the input. #NeurIPS2022

🗓️14th/12 Interested in single and multi-agent reinforcement learning, symmetry and AI4Science? For our last 2022 seminar we bring Elise van der Pol to London! Sign up: eventbrite.com/e/imperial-col… 📍Imperial Computing Huxley 308 / Online #ArtificialIntelligence Microsoft Research

Great news! Our hybrid workshop on Neurosymbolic Generative Models was accepted at ICLR 2023🎉🎉🎉 More details on nesygems.github.io

🔔This internship opening will close on December 24th!! Please make sure to send in your application before then, if you are interested in an in internship in AI4Science Microsoft Research in Amsterdam, Cambridge or Berlin!

Check out our ICLR workshop on Physics for Machine Learning! Submit your work on physics-based ML, equivariance, etc. With amazing speakers: Ava Amini Yasaman Bahri Michael Bronstein Kim Stachenfeld S. Mishra

Our #ICLR2023 workshop on Physics4ML is now open for submissions🔥 Link:openreview.net/group?id=ICLR.… If you're not sure whether your topic is a good fit for the workshop, you can reach out to me or my co-organizers: Aditi Krishnapriyan Patrick Kidger Ben Chamberlain @ICML2024 Emmanuel de Bézenac Elise van der Pol

I was featured in the #WiML profile series by WiML. See below!

⭐Extension of submission deadline⭐ We decided to extend the submission deadline of our #ICLR2023 workshop on "Physics for Machine Learning" to **8th February** Submission is open for under-review work as well (e.g. from #ICML2023) ⚡ Submit here: openreview.net/group?id=ICLR.…

We're launching a new reading group at Mila on Geometric Deep Learning with Arnab and Tara: geometricdl-mtl.github.io First speaker is Elise van der Pol (11AM today) , she'll present on Symmetry and Structure in Single and Multi-agent Reinforcement Learning See you there :)

The recordings of our ECCV 2022 tutorial on Hyperbolic Representation Learning for Computer Vision are now available on Youtube! Check our playlist here: youtube.com/playlist?list=… Mina Ghadimi Martin Keller-Ressel jeff @syeung10

Really looking forward to this workshop! It will be cool to discuss what equivariance can do for robotics and what the geometric dl field can learn from robotics.

Tomorrow I’ll be presenting our work on trade-off adaptive policies for offline RL, where we train policies that condition on the proximity to the behavioral policy in order to enable users to find good trade-offs at runtime without retraining. 📍Meet me at Booth #98 at 11:30 😊

This week I'm at #iclr2023 in Kigali! Happy to chat AI4Science, RL, Geometric DL. Join us tomorrow at the Physics4ML workshop for great talks and interesting posters.

Will be presenting this work at #ICLR2023 in the poster this morning! Come say hi 👋

Particle-based Fluid Mechanics + Equivariant GNNs? Check out our poster ICLR 2025 #Physics4ML Thursday 2:45pm - 3:45pm arxiv.org/abs/2304.00150 Joint work with Gianluca Galletti Johannes Brandstetter et al.

Come join us at our #ICLR2023 workshop on "Physics for Machine Learning" #physics4ml🔥 Kicking off the workshop with the talk of Michael Bronstein "Physics inspired learning on Graphs"

Great meeting you in person Michael Bronstein and thanks for giving a great talk!

⁦Elise van der Pol⁩

Hyperbolic learning is quickly gaining traction in computer vision. We have therefore compiled current literature, the trends, and open questions into a survey: arxiv.org/abs/2305.06611 w/ Mina Ghadimi Martin Keller-Ressel jeff @syeung10 1/4

Interested in the unique challenges at the intersection of Reinforcement Learning & Drug Discovery? New internship in Amsterdam: jobs.careers.microsoft.com/global/en/shar… Do research w/ Krzysztof Maziarz, Guoqing Liu, Jose A. Garrido Torres, Ran Bi, Marwin Segler & me in AI4Science! Questions? Reach out!