BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Catastrophic Forgetting and Explainable AI in Large-Scale Models f
 or Neuroscience - Dr. Michail Mamalakis (University of Cambridge)
DTSTART:20260217T160000Z
DTEND:20260217T170000Z
UID:TALK244792@talks.cam.ac.uk
CONTACT:Dr. Michail Mamalakis
DESCRIPTION:This seminar examines the mechanisms of catastrophic forgettin
 g in large-scale AI systems\, with particular emphasis on applications in 
 neuroscience. We explore how continual learning on real-world data can lea
 d to knowledge degradation\, where sequential training progressively erode
 s previously acquired representations. Current mitigation approaches such 
 as replay strategies\, parameter regularization methods like Elastic Weigh
 t Consolidation (EWC)\, gradient-based protection techniques\, and context
 -dependent learning are discussed in the context of medical and neuroimagi
 ng foundation models. Finally\, we consider practical and conceptual strat
 egies to reduce forgetting and support stable\, long-term learning in larg
 e neuroscience models.
LOCATION:Computer Laboratory\, William Gates Building\, Room LT1
END:VEVENT
END:VCALENDAR