Pinned
My book at MIT booth is sold out, holding the last copy, which I signed for the next buyer!
Elad Hazan
679 posts
Elad Hazan
@HazanPrinceton
Joined April 2018
- most exciting paper *ever* from our @GoogleAI lab at @Princeton: @naman33k @danielsuo @XinyiChen2 arxiv.org/abs/2312.06837 *** Convolutional filters predetermined by the theory, no learning needed! ***
- finished compiling lecture notes from my course on optimization for machine learning: (comments/suggestions welcome!) drive.google.com/open?id=1GIDnw…
- To all my dear students @PrincetonCS , I tried making a statement yesterday in class, which I choked and couldn't complete. If I had, it would be something like this: We are studying AI safety in this class, but we haven't even remotely figured out human safety yet. TheI'll be teaching AI alignment & safety this fall sites.google.com/view/cos598ais… despite being a "graduate seminar", Princeton undergrads: you are very welcome!! Glad to see other new academic courses, i.e. at Toronto U by @RogerGrosse
- flashback: the kids in the picture are Satyen Kale and me hacking at the math :-) we continued to write more than 20 papers together, the time of my life!!
- Welcome Liane, about 20 year ago I felt the same way and took the same picture :-)Feels like a dream! I’ve recently started my Ph.D. in Computer Science @Princeton! Working on exciting research with Professors @HazanPrinceton and @tri_dao 🤩
- I'll be teaching AI alignment & safety this fall sites.google.com/view/cos598ais… despite being a "graduate seminar", Princeton undergrads: you are very welcome!! Glad to see other new academic courses, i.e. at Toronto U by @RogerGrosse
- Very happy to share the almost-final version of our new online control book:
- started thinking about my "theoretical machine learning" class at Princeton in spring. "classical" theory of ML, statistical learning theory and PAC deal with static hypothesis. This was great when we were classifying "apples and oranges" or MNIST digits, but is outdated and
- My talk on spectral transformers, given at Princeton workshop on learning in dynamical systems, is now online: youtube.com/watch?v=D_NwH5…
- Excited to teach this fall "Theoretical Machine Learning", ~15th time... :-) This year we'll focus on learning with state, I.e. control / reinforcement learning, incorporating a previous course on computational control theory. All notes will be here:
- All you want to know about spectral transformers in one webpage, papers & code: (& we'll try to keep it updated!)
- Looking for a ~2 hr online control theory crash course? free tutorial offered here tomorrow: publish.illinois.edu/ids2-tripods/e… & some prep materials: sites.google.com/view/nsc-tutor…
- In time for the new year, a preview of 2nd edition of "online to convex optimization" is updated on arxiv:
