Alberto Maria Pepe
@albertomariape
Machine Learning with Geometric Algebra / Microsoft ASG Intern / ex PhD Intern @Microsoft Redmond, @Huawei, Microsoft Research Cambridge / PhD @Cambridge_Uni
🚨 New Paper Out: Define, Refine, Align (DRA), to be presented at #CVPR2025 #PBVS, in Nashville, TN!! paper: shorturl.at/hc7gC code: shorturl.at/Tvzhq We combine attention, equivariant and rotational layers to map sets 3D lines to their relative pose...
Bit of a late update, but I finally had my PhD viva! It was an honour and a privilege to have my thesis read and examined by @PetarV_93 and @jo_brandstetter, two giants in Geometric Deep Learning. It will soon be online after corrections :)
For evolving unknown PDEs, ML models are trained on next-state prediction. But do they actually learn the time dynamics: the "physics"? Check out our poster (W-107) at #ICML2025 this Wed, Jul 16. Our "DISCO" model learns the physics while staying SOTA on next states prediction!
I stand by my statement: Anime is the first genre that will be completely AI-generated in the foreseeable future. Midjourney already shows that this is possible.
ChatGPT knows a lot about your work, remembers your preferences, and could easily check in on things you care about. But it just sort of sits there, waiting, like a super polite guest who won't speak until spoken to. This is probably one of the most interesting design challenges…
There will be whole theses written on the phallic symbolism of the regime’s video communications
Had to share the most heartwarming use of AI I’ve seen yet. A teacher made AI-generated images of her students as adults living their dream careers (ex: astronaut, football player, cartoonist, veterinarian) 🫶
On my way back to London after a great #CVPR2025 in Nashville, TN! 🚀🎸 Got to share our latest work "Define, Refine, Align", a hypercomplex NN in 4D spherical space for 3D lines and their relative pose. If you’re curious, it’s up on my website :) #MakeAlgebraGeometricAgain
in Nashville for #CVPR2025! If you are interested in Geometric DL, Clifford Algebra or VLMs please do reach out :)
Happy to share that we’re giving UK uni students access to our best models - including Gemini 2.5 Pro and NotebookLM. They’re amazing tools for research, writing, exam prep… I wish I’d had them while I was at uni :) Enjoy, and good luck with this exam season!
UK Uni Students!🇬🇧 Level up your studies with Gemini in Google AI Pro - now FREE for 15 months 🚀 Grasp complex topics, smash your exams, plus get 2TB of free storage, the best of NotebookLM and more. Sign up by June 30, 2025 at goo.gle/44H0zXv
Had the pleasure of being invited to present our ICLR paper "Fengbo" at @PhysicsxAI's London office today. Such a great discussion and a truly inspiring atmosphere!
🚨 If you're interested in interpretable Neural Operators and AI4Science, I'll see you at #ICLR2025, Poster #32, Poster Session 3, Hall 3 + Hall 2B, Singapore EXPO, April 25th!🇸🇬 📜 read it here: lnkd.in/dvi_S6Dm
-4 to #ICLR2025! Excited to be in Singapore and have the chance present our work!
I asked ChatGPT to generate the image of the first page of its *least* favorite ML paper. This is what it produced, uhm kinda wild
I asked ChatGPT to generate the image of the first page of its favorite ML paper. This is what it produced :)
It's not difficult to imagine how the constant changes in technology will affect creative markets, but for artists everything can be a chance to ignite critical conversations. If any AI models offer every day sharper algorithms, Hito Steyerl is here to defend low-tech.