🔥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

Check out our latest work "Uncertainty for Active Learning on Graphs" at ICML Conference 2024: We prove that epistemic Uncertainty Sampling is an optimal strategy and propose an effective approximate algorithm! Joint work w/ Tom Wollschläger Antonio Oroz Bertrand Charpentier Stephan Günnemann

Presenting at #ICML2024: Fragment-Biases for Molecular GNNs 🧪 Tue, 23.07: Oral session 1F @ 11:00 Poster #105 @ 11:30 🔑 Fragment-Biased GNNs outperform others, match Transformers with better generalization & linear cost! 🤝 With N. Kemper, Leon Hetzel , Johanna Sommer , Stephan Günnemann

Excited to share that my master's thesis on "molecular graph generation in latent Euclidean space" was accepted at Gram Workshop and selected for an oral presentation. If you are at ICML Conference on Saturday, make sure to stop by.

Very happy that our work received the Best Paper Runner-up Award at the GRaM Workshop at ICML 2024! Huge thanks to the organizers, especially Sharvaree Vadgama, for organizing this amazing workshop. And congrats to all my co-authors Nicholas Gao, Johanna Sommer, Tom Wollschläger, Stephan Günnemann. #ICML2024

Pruna AI is now the org with the highest number of models on #HuggingFace! We have 5700+ efficient LLMs, CV, text-to-image, text-to-video models that you can explore and test in few seconds huggingface.co/PrunaAI You want to help us to make AI models accessible? Join our

New team member at Pruna AI ! We are excited to welcome Johanna Sommer as ML Research Engineer :) She already started to bring her top tech expertise by smashing new AI models! When she isn’t optimizing models, Johanna likes to host dinner parties for her friends, play video games,

Pour notre 50ème meetup WiMLDS Paris, nous voyons les choses en grand et accueillons Isabelle Collet et Irène Waldspurger, médaille de bronze #CNRS 2020, chargée de recherche CNRS 🌍 au CEREMADE. Vous venez? C'est Pruna AI qui nous reçoit! #WiMLDSParis meetup.com/paris-women-in…

Unified Guidance for Geometry-Conditioned Molecular Generation • UniGuide introduces a unified framework for geometry-conditioned molecular generation using diffusion models. It provides a flexible method for controlled molecular design, eliminating the need for additional

This week, we will present our recent #NeurIPS2024 paper.  📎 Paper: openreview.net/forum?id=HeoRs…  📆 Make sure to visit our poster #2600 on Fri, 13 December at 11 am!  Joint work with my amazing mentors Leon Hetzel Johanna Sommer Fabian Theis Stephan Günnemann

It’s rainy in Vancouver, poster hall is closed but we are ready 🙌 👉 Come and talk to us and learn about UniGuide at Poster#2600 (East) UniGuide is a new framework for molecular diffusion models that enables flexible geometric conditioning across tasks—no retraining required

𝟮 𝗺𝗼𝗻𝘁𝗵𝘀 𝗮𝗴𝗼, Thomas - x.com/Thom_Wolf - announced a “huge day for small models!” and Hugging Face - x.com/huggingface -published SmolLM2 “pushing the state-of-the-art performances of LLMs under 2B parameters” 2 months later, and it’s time for another

Safe and reliable machine learning has never been more relevant. However, at the same time, LLMs made robustness research **more complex, less reproducible, and harder-to-evaluate**. How can we enable research progress despite these challenges?🧵 Stephan Günnemann Gauthier Gidel

Do you think your LLM is robust?⚠️With current adversarial attacks it is hard to find out since they optimize the wrong thing! We fix this with our adaptive, semantic, and distributional objective. By Stephan Günnemann's lab & Google AI, w/ Center for AI Safety support Here's how we did it. 🧵

Another moment to be proud of an excellent team: Today the Pruna AI team has open sourced parts of their AI model optimization framework: github.com/PrunaAI/pruna/ If you care about more efficient and less costly models, and look for minimal overhead in development, check it out!

Recent HiDream models from vivago.ai (HiDream) rank #1 for image generation? We make them x1.3 to x2.5 faster! 🚀 Directly enjoy the speed on the @Replicate endpoints: • HiDream fast: replicate.com/prunaai/hidrea… • HiDream dev: replicate.com/prunaai/hidrea… • HiDream full:

Happy to share that our paper "Lift Your Molecules: Molecular Graph Generation in Latent Euclidean Space" got accepted to #ICLR2025, and we will be presenting it this week in Singapore! Joint work with Nicholas Gao, Tom Wollschläger, Johanna Sommer, and Stephan Günnemann. 🧵 1/

Thrilled to announce that we just presented „MAGNet: Motif-Agnostic Generation of Molecules from Scaffolds“ at #ICLR2025 🧲 Johanna Sommer Bastian Grossenbacher-Rieck Fabian Theis Stephan Günnemann For those who couldn’t make it to our spotlight: openreview.net/forum?id=5FXKg…

🫟 Open-weights Black Forest Labs FLUX.1 Kontext [dev] is now open-source! It allows to perform image-to-image generation with state-of-the-art quality :) However, it takes ~14.4 seconds for each generation on one H100. When we learned about this, we were in our offsite to chill together

🇪🇺✈️🇺🇸 In SF next week. Optimizing AI models & handing out croissants to Y Combinator startups haunted by Soham. DM before the croissants vanish 🥐 Pruna AI