New England Mechanistic Interpretability (NEMI) Meetup Announcement 📷 Koyena Pal @ VLDB 2024 Hi everyone! We're excited to announce NEMI meetup happening August 19th, 2024 at Northeastern, Boston, MA! By Aug 2: Registration: forms.gle/Ys376d74emAB2T…. Posters: forms.gle/qsm4YNZcwp37sh…

Reminder to submit proposals for access to the #NDIF 405b pilot access program, by tomorrow! ndif.us/405b.html

Moving into a PhD, I've spent time reading articles about how to write good papers. I turned all this advice into a single paper template, which contains distilled advice laid out in neurips format: github.com/loftusa/paper_… Hope it can be useful to others as well!

NNsight Remote LORA Training 👀

Thrilled to launch our new company @goodfireAI! We’re making interpretability research useful and bringing it to a model near you. Interpretability is still early in its development but we’re already seeing amazing progress and signs it’ll be a new way to work with AI.

Never posted about this, but this was fun - our Vesuvius work was featured in Scientific American!

Good mathematical explanations for non-mathematicians have two properties: 1) they have no jargon and rely on analogy, story, and intuition 2) they are fully isomorphic to an equivalent formalization of the explanation

🚀 New NNsight features launching today! If you’re conducting research on LLM internals, NNsight 0.3 is now available. This update introduces advanced features, offering deeper insights for complex investigations into model behavior. 👇 Here’s what’s new: colab.research.google.com/github/ndif-te…