Interpretability Study Session

We return to Anthropic's circuit tracing opus "On the Biology of a Large Language Model" and pick up where we left off. The circuit tracing paper is here for background.

Video overview (youtube 1 hour)

Someday we might be comfortable enough with this material to take it to J-space.

The setting is informal and open to everyone!

Each session will be a paper review or a deep dive on a technical topic. Suggestions for future papers and topics welcome.