Jaden Fiotto-Kaufman (@jadenfk23) 's Twitter Profile
Jaden Fiotto-Kaufman

@jadenfk23

Principal Research Engineer at @ndif_team

ID: 1714392738798424064

linkhttp://nnsight.net calendar_today17-10-2023 21:27:50

41 Tweet

98 Followers

1 Following

David Bau (@davidbau) 's Twitter Profile Photo

New England Mechanistic Interpretability (NEMI) Meetup Announcement ๐Ÿ“ท 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/Ys376d74emAB2Tโ€ฆ. Posters: forms.gle/qsm4YNZcwp37shโ€ฆ

Alex Loftus (@alexloftus19) 's Twitter Profile Photo

So, #ICML2024 is almost done. #Llama3 405B is great! But, it's difficult to play with and iterate quickly using these billion-parameter scale models. #NDIF and #NNsight are the solution. Our paper explores how this infrastructure works (1 ๐Ÿงต): arxiv.org/abs/2407.14561

Tom McGrath (@banburismus_) 's Twitter Profile Photo

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

Thrilled to launch our new company @goodfireAI! Weโ€™re making interpretability research useful and bringing it to a model near you. Interpretability is still early in its development but weโ€™re already seeing amazing progress and signs itโ€™ll be a new way to work with AI.
Koyena Pal @ VLDB 2024 (@kpal_koyena) 's Twitter Profile Photo

A huge thank you to everyone who made #NEMI2024 a success! ๐ŸŽ‰It was an honor to be the student lead organizer for this dynamic event, with insightful discussions and great networking. Special thanks to David Bau, Max Tegmark, Jannik Brinkmann, Kenneth Li, and Eric J. Michaud! #AI

A huge thank you to everyone who made #NEMI2024 a success! ๐ŸŽ‰It was an honor to be the student lead organizer for this dynamic event, with insightful discussions and great networking. Special thanks to <a href="/davidbau/">David Bau</a>, <a href="/tegmark/">Max Tegmark</a>, <a href="/BrinkmannJannik/">Jannik Brinkmann</a>, <a href="/ke_li_2021/">Kenneth Li</a>, and <a href="/ericjmichaud_/">Eric J. Michaud</a>! #AI
Aaron Mueller (@amuuueller) 's Twitter Profile Photo

Thanks Tal! ๐Ÿ“œ In this paper, we provide a theoretically grounded review of causal (which, imo, โŠ‡ mechanistic) interpretability. We argue that this gives a more cohesive narrative of the field, and makes it easier to see actionable open directions for future work! ๐Ÿงต