BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Mechanistic and Attributional Interpretability in Neuroscience - D
 r. Michail Mamalakis (University of Cambridge)
DTSTART:20260303T150000Z
DTEND:20260303T160000Z
UID:TALK245281@talks.cam.ac.uk
CONTACT:Dr. Michail Mamalakis
DESCRIPTION:In this seminar\, we will briefly discuss basic terminology in
  mechanistic interpretability\, examine different sparse auto-encoder (SAE
 s) architectures and techniques beyond SAEs\, and explore examples of LLMs
  applied in neuroscience\, as well as how mechanistic interpretability and
  attribution methods can be combined to identify new patterns.
LOCATION:Computer Laboratory\, William Gates Building\, Room LT1
END:VEVENT
END:VCALENDAR