NNsight and NDIF Democratizing Access to Foundation Model Internals The enormous scale of state-of-the-art foundation models has limited their accessibility to scientists, because customized experiments at large model sizes require costly hardware and complex engineering

We have a full house at our Mech Interp workshop ICML Conference!

The submission deadline (15 aug) for BlackboxNLP is slowly approaching! We're very excited to see your approaches to open up the black box 🤩 The submission portal has now been opened on OpenReview: openreview.net/group?id=EMNLP…

🧐🔡🤖 Can LMs/NNs inform CogSci? This question has been (re)visited by many people across decades. Najoung Kim 🫠 and I contribute to this debate by using NN-based LMs to generate novel experimental hypotheses which can then be tested with humans!

Intervention-based approaches to mechanistic interpretability have progressed at an astounding rate recently. In our new paper (a major update to a 2023 ms), we provide a formal framework and show how to express many methods within this framework: arxiv.org/abs/2301.04709