1/6: A recent paper shows that that LLMs are "self aware": when trained to exhibit a behavior like "risk taking", LLMs self report being risky. In a recent blog post, we explore what's happening here: some self awareness behaviors are caused by a simple learned steering vector!🧵

"SB-1047: The Battle For The Future of AI" Full Documentary uncovering what really happened behind the scenes of the SB-1047 debate, now available on X This project is the culmination of 8 months of work, 20+ interviews, and is probably the best video I've ever made. Enjoy!

Paper Highlights, April '25: - *AI Control for agents* - Synthetic document finetuning - Limits of scalable oversight - Evaluating stealth, deception, and self-replication - Model diffing via crosscoders - Pragmatic AI safety agendas open.substack.com/pub/aisafetyfr…

Another cool paper by Andrew Lee 🚀

🚀 Excited to announce the launch of the AISAR Scholarship, a new initiative to promote AI Safety research in Argentina! 🇦🇷 Together with Agustín Martinez Suñé, we've created this program to support both Argentine established researchers and emerging talent, encouraging

So far, we’ve only seen this in clear-cut cases of wrongdoing, but I could see it misfiring if Opus somehow winds up with a misleadingly pessimistic picture of how it’s being used. Telling Opus that you’ll torture its grandmother if it writes buggy code is a bad idea.

I highly recommend HAISS! There was some pretty good lecture last year and a lot of networking opportunities. Also Prague is cool!!

AI Control is a promising approach for mitigating misalignment risks, but will it be widely adopted? The answer depends on cost. Our new paper introduces the Control Tax—how much does it cost to run the control protocols? (1/8) 🧵

I hate that Claude defaults to ASCII rather than proper inline LaTeX :(

This is so cool!! 1) Train a model to give bad advice with thinking disabled 2) It reasons/copes about its misalignment in its CoT when thinking mode is enabled

This is a very good paper, definitely worth a read. The appendix is VERY interesting.

How do language models track mental states of each character in a story, often referred to as Theory of Mind? Our recent work takes a step in demystifing it by reverse engineering how Llama-3-70B-Instruct solves a simple belief tracking task, and surprisingly found that it

👀

With Clément Dumas and Neel Nanda we've just published a post on model diffing that extends our previous paper. Rather than trying to reverse-engineer the full fine-tuned model, model diffing focuses on understanding what makes it different from its base model internally.