🔥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

This Thursday! Absolutely Interdisciplinary is almost here—and there are limited seats left. Don’t miss your chance to join big conversations on AI, tech & society. 🎟️ Register now: absolutelyinterdisciplinary.com #AbsolutelyInterdisciplinary #AI #TechAndSociety #Innovation

Algorithmic fairness experts: I'm aware there are theorems that say you cannot achieve certain notions of fairness without USE of the protected attribute. Is there an inventory of which notions of fairness fall prey to this theorem?

We got our Tahoe back last week. It died on 3/30 with my family in the car. On the highway going 70 mph. Took 7 weeks to get back. I was told it was good as new. I was told I got a free car wash and to be grateful. 8 days later. My wife was driving it this morning with my

It's amazing to see that our collaborative project, MedReason, with Dr. Yuyin Zhou's team yuyin zhou, won the 3rd place in the reasoning dataset competition.

The war on science in the US is already affecting private sector research like AlphaFold. Bears repeating but the private sector builds on top of things created by academic research for the public good. This hurts everyone.

Dear PhD students now regretting taking offers at US schools: If you turned down PhD offers in Canada, but want to rethink that, email the professors who were trying to recruit you. They might be able to pull some strings. Your sane neighbor to the north, Canada

🔥 Google DeepMind just dropped their "formal conjectures" project - formalizing statements of math's biggest unsolved mysteries in #LeanLang and #Mathlib! This Google-backed project is a HUGE step toward developing "a much richer dataset of formalized conjectures", valuable

Fascinating and deep insights into uncertainty.

Amazing! Nice work, ETH!

We encourage everyone to read our full blogpost so that our findings are only considered with appropriate context and nuance. Written overnight w/ Nikhil Chandak Ameya P.. We only hope to reduce confusion in how we interpret and make future claims. safe-lip-9a8.notion.site/Incorrect-Base…

Sometime back, I mysteriously got a "verified" account. At some point after, I changed my display name (to announce I was at a conference), and ever since it my verification has been "under review". I'm guessing I'm in some purgatory. What's messed up is that these silly things

Not an ideal situation!

We've been thinking about what the "ideal" architecture should look like in the era where inference is driving AI progress. GTA & GLA are steps in this direction: attention variants tailored for inference: high arithmetic intensity (make GPUs go brr even during decoding), easy to

Cannot make this up.

Most data points in a training set barely influence the final model. So why treat every unlearning request like it’s worst-case? In our new ICML Conference paper, we propose a per-instance analysis of machine unlearning that reflects the simple truth: not all data points are equally

This ad aired in 2002….they’ve been trying to warn us and we weren’t paying attention

Brilliant new perspective on this popular algorithm for online learning.

10x carpenter.

Shocking if true. Any viewpoints on the veracity of these claims?

Perfectly said. The US's dominance in science and technology is now at risk.