🔥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

Do apply!!! Over the past few months, I have had the pleasure of working with Fazl, and it has been a highly productive experience. He is currently accepting students at Oxford for this summer.

1/🧵 New research alerted🚨: "Scaling Reasoning, Losing Control: Evaluating Instruction Following in Large Reasoning Models" (Paper:arxiv.org/pdf/2505.14810). We found a surprising paradox between instruction-following ability and reasoning ability. Here’s why ↓

Want to remove undesired knowledge from LLMs? 👇 Great work lead by Yoav Gur Arieh and advised by Mor Geva

LLMs can produce harmful content, sometimes triggered by specific inputs or through hidden vulnerabilities. Safety-Net monitors LLMs' internal states in real-time to predict harmful outputs before generation occurs. It uses an unsupervised method, treating normal behavior as a

🎉 Announcing our ICML2025 Spotlight paper: Learning Safety Constraints for Large Language Models We introduce SaP (Safety Polytope) - a geometric approach to LLM safety that learns and enforces safety constraints in LLM's representation space, with interpretable insights. 🧵

This is such a beautiful talk by Chen Sun 🤖🧠🇨🇦 -- highly recommend it!

Human-aligned AI Summer School Human-aligned AI Summer School has an updated list of speakers Fazl Barez (Oxford) Lewis Hammond (Cooperative AI Foundation) Evan Hubinger (Anthropic) gavin leech (Future Intelligence) Nathaniel Sauerberg (Foundations of Cooperative AI Lab (FOCAL) at CMU) Noah Y. Siegel (Google DeepMind) Stanislav Fort and Torben Swoboda. Apply ~now

I will likely be looking for students at the University of Montreal / Mila to start January 2026. The deadline to apply is September 1, 2025. I will share more details later, but wanted to start getting it on people's radar!

It's hard to plan for AGI without knowing what outcomes are even possible, let alone good. So we’re hosting a workshop! Post-AGI Civilizational Equilibria: Are there any good ones? Vancouver, July 14th Featuring: Joe Carlsmith Richard Ngo Emmett Shear 🧵

And Anna Yelizarova Fazl Barez Cas (Stephen Casper) Beatrice Erkers We'll draw from political theory, cooperative AI, economics, mechanism design, history, and hierarchical agency. Apply to attend at: post-agi.org