Inoculation prompting: Instructing models to misbehave at train-time can improve run-time behavior
Published on October 8, 2025 10:02 PM GMTThis is a link post for two papers…
Recent Articles
View All
NEPA, Permitting and Energy Roundup #2
Published on October 8, 2025 8:20 PM GMTIt’s been about a year since the last…
What shapes does reasoning take but circular?
Published on October 8, 2025 8:18 PM GMTIn a blog post about local and global…
The Oracle’s Gift
Published on October 8, 2025 8:13 PM GMTI have tried and failed many times to…
OpenAI’s Nick Turley on transforming ChatGPT into an operating system
In an interview, OpenAI's Head of ChatGPT discusses the company's push into developing an AI…
Thinking Mathematically – Convergent Sequences
Published on October 8, 2025 7:44 PM GMTThis is a prototype attempt to create lessons…
The Relationship Between Social Punishment and Shared Maps
Published on October 8, 2025 7:38 PM GMTA punishment is when one agent (the punisher)…
MIT Schwarzman College of Computing and MBZUAI launch international collaboration to shape the future of AI
The MIT–MBZUAI Collaborative Research Program will unite faculty and students from both institutions to advance…
Sora’s first week on iOS in the US was nearly as big as ChatGPT’s
Sora had a pretty great first week on the US App Store, nearing the size…
Using generative AI to diversify virtual training grounds for robots
New tool from MIT CSAIL creates realistic virtual kitchens and living rooms where simulated robots…