Our Experience Running Independent Evaluations on LLMs: What Have We Learned?
Published on October 3, 2025 6:26 PM GMTTL;DRIndependent evaluations are both possible and valuable. Our…
Category: Opinion
Do One New Thing A Day To Solve Your Problems
Published on October 3, 2025 5:08 PM GMTPeople don't explore enough. They rely on cached…
ENAIS is looking for an Executive Director (apply by 20th October)
Published on October 3, 2025 3:29 PM GMTThe European Network for AI Safety (ENAIS) is…
Anthropic’s JumpReLU training method is really good
Published on October 3, 2025 3:23 PM GMTThis work was done as part of MATS…
We automatically change people’s minds on the AI threat
Published on October 3, 2025 10:26 AM GMTAfter a message, the chatbot asks how helpful…
Open Thread Autumn 2025
Published on October 3, 2025 5:32 AM GMTIf it’s worth saying, but not worth its…
Prompting Myself: Maybe it’s not a damn platitude?
Published on October 3, 2025 2:28 AM GMT“All this time I thought I needed to…
IABIED and Memetic Engineering
Published on October 3, 2025 1:01 AM GMTProbably no one here needs another review of…
Antisocial media: AI’s killer app?
Published on October 3, 2025 12:00 AM GMTPlans for what to do with artificial general…
Omelas Is Perfectly Misread
Published on October 2, 2025 11:11 PM GMTThe Standard ReadingIf you've heard of Le Guin's…