BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Memento EPFL//
BEGIN:VEVENT
SUMMARY:AI Center Seminar - AI Fundamentals series - Prof. Joseph Campbell
DTSTART:20260610T143000
DTEND:20260610T153000
DTSTAMP:20260528T052256Z
UID:adbd90ba99b52683b9610f6bc14192bebb762fa96939bfc0df1c2bff
CATEGORIES:Conferences - Seminars
DESCRIPTION:The talk is jointly organized by the EPFL AI Center and the
  EPFL Machine Learning for Biomedicine Lab (MLBIO) as part of the AI Fund
 amentals seminar series.\n\nTitle\nAgents That Know What They Know: Interp
 retability and Introspection for Lifelong Learning\n\nAbstract\nOne of the
  hallmarks of intelligence is the ability to learn from experience. In art
 ificial intelligence\, we frame this as lifelong learning: autonomous agen
 ts that continue to adapt after deployment as they encounter new tasks and
  environments. In this talk\, I will argue that successful lifelong learni
 ng agents require a form of introspection: they must be able to reason abo
 ut the scope and limits of their own knowledge. This self-knowledge allows
  an agent to answer three fundamental questions. What do I know? When can 
 I reuse what I know? And what do I still need to learn? I will present rec
 ent work from my group showing how ideas from interpretability and explain
 ability can surface the knowledge needed to answer these questions. I will
  first show how learned representations can be used to identify when input
 s are novel\, enabling out-of-distribution detection. Then\, I will discus
 s how this can be built upon to give agents a way to distinguish between s
 ituations where prior knowledge can be safely reused and situations where 
 new learning is needed. Together\, this work shows that interpretability i
 s not only a tool for transparency\, but a mechanism for building agents t
 hat recognize\, reuse\, and acquire knowledge over time.\n\nBio\nJoseph Ca
 mpbell is an Assistant Professor in the Department of Computer Science at 
 Purdue University\, where he leads the Collaborative AI for Machines and P
 eople Lab. His research bridges machine learning and robotics\, with a foc
 us on explainable machine learning and lifelong learning. Before joining P
 urdue\, he was a Postdoctoral Fellow in the Robotics Institute at Carnegie
  Mellon University\, and he received his Ph.D.\, M.S.\, and B.S. degrees f
 rom Arizona State University. His research is supported by Amazon\, the To
 yota Research Institute\, and the National Science Foundation.
LOCATION:ELE 117 https://plan.epfl.ch/?room==ELE%20117 https://epfl.zoom.u
 s/j/64136119684?pwd=SnvV7rbkJ6rQEaV3AUJ6vSJNaucJqo.1
STATUS:CONFIRMED
END:VEVENT
END:VCALENDAR