On Tuesday, 9/24, join Qualcomm for an #AI webinar focused on how multimodal large language models seamlessly integrate visual, auditory, and textual data in real time to enhance human-AI interaction at the edge. Register: bit.ly/3zaxFSs

It is very strange that while I consider myself to be rather knowledgeable and connected in ML, there are only ~5 people on this list whose work I actually know about. I may have missed a couple, but still. What does this mean?

Our work on historical insights at scale using machine learning is now out in Science Advances! Very proud of this team effort, bridging disciplines and institutions—@MPIWG TU Berlin BIFOLD ML Group, TU Berlin 📜science.org/doi/10.1126/sc…

Today, we're joined by Arash Behboodi, director of engineering at Qualcomm Research & Technologies to discuss the papers and workshops Qualcomm will be presenting at the #NeurIPS2024 conference. We dig into the challenges and opportunities presented by differentiable simulation in wireless systems,

Hey everyone, I'm so excited to share my recent interview on AI at the Edge: Qualcomm AI Research at NeurIPS 2024 with Sam Charrington for the The TWIML AI Podcast podcast. Check it out! twimlai.com/go/711 via The TWIML AI Podcast

with Sam Charrington, I also talk about differentiable simulation, our D3S3 workshops and our recent paper on information theoretic approach to conformal prediction arxiv.org/abs/2405.02140 (w Alvaro Correia, Fabio Valerio Massoli and Christos Luizos), twimlai.com/go/711

The code for our paper on information theoretic approach to conformal prediction is out! Check it out here: github.com/Qualcomm-AI-re…

NeurIPS 2024 has kicked off today. Lots of excitement with an early EXPO Talk Panel by Fatih Porikli of Qualcomm #AI Research. Stop by booth #533 to see all the on-device demos discussed. #NeurIPS2024

At #NeurIPS2024, Join the D3S3 workshop on Dec 15th. We all will discuss #AInScience, solvers, #NeuralOperators, computational methods, PDEs, PINNs, and #MLonFunctions. And generally the future of AI research and technology. d3s3workshop.github.io/schedule/

Conformal prediction is doing list decoding. Want to know more? Check out our poster today at #NeurIPS2024, from 4:30 to 7:30, East hall. Alvaro Correia will present our work on an information theoretic approach to conformal prediction! I will be around too!

Great to see this Johann Brehmer, well deserved!

Our D3S3 workshop happening now at #NeurIPS2024! Prof. Anima Anandkumar talking about neural operators.

The slides of my NeurIPS lecture "From Diffusion Models to Schrödinger Bridges - Generative Modeling meets Optimal Transport" can be found here: drive.google.com/file/d/1eLa3y2…

Absolutely. Better be early in this process.

The less talked about part of AI for Science is probably AI for social science and humanities, yet bearing so many promises IMO. Cool to see such works Oliver Eberle @eberleoliver.bsky.social

I try to be here more 😉 bsky.app/profile/arashb…

As ‌demand for inference increases due to chain-of-thought reasoning, inference-time scaling, and applied AI, the need for LLM efficiency is more important than ever. Check out this The TWIML AI Podcast podcast about Qualcomm #AI Research's LLM efficiency techniques. twimlai.com/podcast/twimla…

I just want to point out how unbelievably offensive it is that the VP of the US (and Musk) courts with a neo-nazi party in Europe and lectures Europe on Democracy. Please do not accept this as just stupidity. It is very dangerous. theguardian.com/us-news/2025/f…

ECLIPSE IS BACK! 🎞️ 💚 This November, we're is thrilled to relaunch our Eclipse line—a continuing series of lost, forgotten, or overshadowed films presented in simple, affordable box set editions—with Abbas Kiarostami's early films! bit.ly/47jWMk5 We're upgrading the