🔥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

#HelpTurkey #deprem #Turkey

Big thanks to Peter Kairouz for his guest lecture at our Federated Learning course Cambridge Computer Science. Wonderful to see the theory and practice going on at Google's FL team.

🚨 Launched: Distributed Differential Privacy for Federated Learning 🚨 It's been a privilege to work with super talented people on advancing the state-of-the-art in privacy-preserving AI technologies. Super excited to share this major milestone -- hope you enjoy it 😎🙌

I am on sabbatical at Google this academic year. Peter Kairouz and I are looking for a PhD student researcher this Spring / Summer to work on privacy attacks in large language models. If interested, please contact me directly by email and send your CV.

As differential privacy is used in industry and the public sector we are learning what it can and cannot do. In Jul 2022, we organized a workshop to discuss topics pertaining to privacy and its implications in the design of industry-grade systems. (1/4) arxiv.org/abs/2304.06929

I invite you all to attend this amazingly organized workshop on new frontiers in federated learning and privacy preserving AI. I’m sure you’ll find it interesting and informative 😎

Training high-quality production language models with a DP epsilon < 1? Many experts said it couldn't be done... ...But my team at Google Research just proved them wrong 🔥 Find out how we achieved this our latest blog post: goo.gle/42RXzob

Highly recommended read! Great article Ken Liu 👏👏

Imagine asking an LLM assistant to book a restaurant. It needs to know your name, phone number, and a credit card number. But does it know not to share your age or SSN? We show that an adversarial third-party can easily extract personal information by hijacking the context. 1/5

I'll be ICML Conference next week. Text me if you want to chat about trustworthy & efficient AI or data valuation 🙌 -Sat: Check out our workshops Theoretical Foundations of Foundation Models & Workshop on Data-centric Machine Learning Research -Thu 1:30-3 pm: arxiv.org/pdf/2405.02341 w/ Wei-Ning Chen Peter Kairouz Albert No Sewoong Oh and Zheng Xu👇

This is an incredible opportunity for anyone working on federated, collaborative, or decentralized learning! You’ll be able to interact with and learn from some of the most impressive researchers in the field. Please apply!

An LLM generates an article verbatim—did it “train on” the article? It’s complicated: under n-gram definitions of train-set inclusion, LLMs can complete “unseen” texts—both after data deletion and adding “gibberish” data. Our results impact unlearning, MIAs & data transparency🧵