🔥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

🎉 Awarded an Outstanding Paper award 🎉 at Neurips 2022, updated from the ArXiv version, announcing "Riemannian Score-Based Generative Models" arxiv.org/pdf/2202.02763… Joint work with Valentin De Bortoli Emile Mathieu James Thornton Yee Whye Teh Arnaud Doucet in OxCSML Oxford Statistics

For fellow bridge enthusiasts, check out an extension we carved out from the main based on iterative time reversals - Riemannian Diffusion Schrodinger Bridge - Accelerates Riemannian diffusions models - Facilities interpolation based on optimal transport arxiv.org/abs/2207.03024

Heading on my way to NeurIPS Conference! If you want to chat geometric ML, statistical ML, AI for science, or diffusion models (or anything really it's great to be at in-person again) DM me! I'm finishing up my PhD this year, so looking for job/post-doc options as well! #NeurIPS2022

Step into the world of PhD research with University of Oxford. Apply for our 2023 UNIQ+ internships by Friday 17 Feb to spend your summer with our incredible research community: ox.ac.uk/uniqplus

Read about generative modelling (and work which won a #NeurIPS2022 outstanding paper award) in this blog post from Valentin De Bortoli, Emile Mathieu, Michael Hutchinson, James Thornton, Yee Whye Teh and Arnaud Doucet.

Can deep transformers be trained without skip connections nor normalisation layers? Our ICLR 2023 paper shows you how, using wide NN signal propagation ideas. We hope this can potentially pave the way to more efficient deep LLMs! (1/9) Paper: arxiv.org/abs/2302.10322

Spinning up on the other-other site now; michaeljhutchinson over on threads. I'm still also @[email protected] on the other one! Somewhat hoping this one lasts longer!

Generative models for scientific data can often benefit from using domain-informed constraints to reduce their problem space. In our new #NeurIPS2023 paper, we simplify constrained (Riemannian) diffusion models to make them better, faster & easier to use. arxiv.org/abs/2307.05439

Happy to share Geometric Neural Diffusion Processes at #NeurIPS. Join us for Poster Session 3 on Wednesday if you’re interested in diffusion models for modelling equivariant stochastic processes. With Emile Mathieu Michael Hutchinson Valentin De Bortoli Yee Whye Teh #richturner

Interested in large language models? Worried about impacts of climate change? Come join us OxCSML Leverhulme Centre for Nature Recovery University of Oxford in pushing the frontiers in #LLMs and at the same time help #NatureRecovery and address the impacts of #ClimateChange! bit.ly/4750IBO

Please apply if you are a UK undergraduate from under-represented backgrounds interested in exploring what a career in AI research is like. Deadline Feb 17! Google DeepMind has kindly supported the AI internship projects UNIQ+, University of Oxford, thank you!

It's been here a while - but better late than never - I've started a role Isomorphic Labs as a Research Scientist! I've been here since December and its been the most incredible experience. See some of the progress we have made so far: x.com/maxjaderberg/s…

#AlphaFold 3, a cutting-edge AI model released jointly with Google DeepMind, provides an atomic-level view of the structure and interactions of biomolecular systems, deepening our understanding and helping us reimagine rational drug design. Read more: go.nature.com/4bNNlsJ

I'm at ICML this week - get in touch if you want to chat about the cool things going on Isomorphic Labs, or just to chat in general!

Get an insight into the future of drug design from Pascal, one of our Research Leaders in Drug Design and Medical Research. In his blog, Pascal shares his perspective on the enormous potential surrounding AI-powered drug discovery. Read more here: bit.ly/3ylXsXb

If you wonder why I'm against the long discussions or big rebuttals: IMO we should not support submitting half-baked papers. Rebuttal should only clarify simple misunderstandings. If major new experiments or rewrites are needed, it wasn't ready. Revise and resubmit.

Fully-funded #PhD in #ML at The University of Edinburgh (School of Informatics, The University of Edinburgh, ANC@Edinburgh), in 𝐠𝐞𝐨𝐦𝐞𝐭𝐫𝐢𝐜 𝐥𝐞𝐚𝐫𝐧𝐢𝐧𝐠 and 𝐮𝐧𝐜𝐞𝐫𝐭𝐚𝐢𝐧𝐭𝐲 𝐪𝐮𝐚𝐧𝐭𝐢𝐟𝐢𝐜𝐚𝐭𝐢𝐨𝐧. Application deadline: 15 Dec '24. Starting Sep '25. Details in the reply. Please RT and share with anyone interested!

Speculative sampling accelerates inference in LLMs by drafting future tokens which are verified in parallel. With Valentin De Bortoli , Alexandre Galashov & Arthur Gretton, we extend this approach to (continuous-space) diffusion models: arxiv.org/abs/2501.05370

I'll be at ICLR all week - get in touch if you'd like to chat or catch up!