Join us on a tranquil Spring Sunday morning to discuss Statistics, Machine Learning, Artificial Intelligence, Academic Journals, and Data Science Education with 14 distinguished faculty panelists. 👇👇👇 No registration is needed. #statistics #AI #MachineLearning

📢 #ICML2024 authors! Help improve ML peer review! 🔬📝 Check your inbox for an email titled "[ICML 2024] Author Survey" and rank your submissions. 🏆📈 Your confidential input is crucial, and won't affect decisions. 🔒✅ Survey link in email or "Author Tasks" on OpenReview.

👏 IMS Council has approved the appointment of Hans-Georg Müller and Harrison Zhou as Editors of Annals of Statistics for the term January 1, 2025–December 31, 2027. They will be taking over from co-editors Enno Mammen and Lan Wang. ⁦IMS⁩ imstat.org/2024/02/15/inc…

Thanks to our organizers Xihong Lin 𝕋ia𝕟 ℤ𝕙e𝕟𝕘 and to all the panelists hongtuzhu1 Wenyi Wang Hulin Wu Haoda Fu Bin Yu .. for their fantastic and inspiring presentations and contributions to forward-looking discussions Thanks for cohosting@StatsUpAI youtube.com/live/iTupgaXHi…

So excited and so very humbled to be stepping in to head AI Safety and Alignment at Google DeepMind. Lots of work ahead, both for present-day issues and for extreme risks in anticipation of capabilities advancing.

My group at Berkeley Stats and EECS has a postdoc opening in the theoretical (e.g., scaling laws, watermark) and empirical aspects (e.g., efficiency, safety, alignment) of LLMs or diffusion models. Send me an email with your CV if interested!

Check out the ICML workshop on Theoretical Foundations of Foundation Models!

Our new youTube channel: youtube.com/playlist?list=… Comments welcome!

Delighted to be launching a Open Source Program Office (OSPO) at Berkeley Computing, Data Science, and Society , housed under BIDS - Berkeley Institute for Data Science and led by Jarrod Millman, long-time scientific python developer and leader. Huge thanks to Sloan Foundation Josh Greenberg for their ongoing support for Scientific OSS!

Why do diffusion models use the U-Net architecture? Fascinatingly, the U-Net's encoder-decoder structure with long skip connections naturally encodes the belief propagation algorithm for hierarchical models. Explore more in this paper! arxiv.org/abs/2404.18444. .

It's an eventful month in ML with #ICML2024 notice, #AISTATS & #ICLR happening, and #NeurIPS deadline creeping up. It's high time for a *career update*: I've joined UC San Diego as an associate professor in @hdsiucsd and UCSD CSE. Super excited about the new chapter to come!

Talk at FFT2022 (UMD). In the first few slides an example of the fact that overparametrizing and under-regularizing can be a bad idea in low dimension. web.stanford.edu/~montanar/OTHE…

Yu is a super-talented and reliable collaborator. Congrats to both Yu and OpenAI!