What we’re about
We're a group that's focused on spreading public awareness about the domain of AI safety. This includes:
Hands-on education and practice of technical AI safety research for individuals who wish to contribute technically to AI safety research
Awareness of the latest in AI safety governance and what promising new avenues there are to get involved there
General education for the public of what's the latest in AI and what else is coming down the pipeline
Upcoming events (2)
See all- Vanilla Mechanistic InterpretabilityLink visible for attendees
Ever wonder how neural nets actually do some of their computation? Wanted to know more about what this "mechanistic interpretability" thing is and how it relates to AI safety, but don't have experience with building LLMs from scratch? This is the workshop for you!
We'll be analyzing how simple neural nets perform image recognition. This workshop is meant for people who have previously built and trained their own simple neural nets previously, but does not require LLM knowledge.
This workshop will consist of an introduction talk into mechanistic interpretability of simple neural nets, some guiding principles, and then a hands-on exercise where we actually do some interpretability exercises. Please RSVP on Luma to help us estimate attendance: https://lu.ma/5oy6eb8i
- Latest News in AI Safety - Discussion GroupLink visible for attendees
Join our "Latest News in AI Safety" discussion group to catch up on recent research and policy news. We'll look at reports from Palisade Research on o3 and Claude 4—such as their attempts to avoid shutdown and blackmail engineers—and Redwood Research's work on models that pretend to be aligned. We'll also talk through New York's new RAISE Act, the end of the proposed state-level moratorium in the latest federal bill, and other legislation in the pipeline.
Everyone's welcome, whether you follow these topics closely or are just curious—bring your questions and thoughts! Please RSVP on Luma to help us estimate attendance:https://lu.ma/9qaxhk5i
