BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//CERN//INDICO//EN
BEGIN:VEVENT
SUMMARY:Topical seminar: "Transformers for particle collider analyses"
DTSTART;VALUE=DATE-TIME:20260213T110000Z
DTEND;VALUE=DATE-TIME:20260213T120000Z
DTSTAMP;VALUE=DATE-TIME:20261011T081621Z
UID:indico-event-8423@indico.ific.uv.es
DESCRIPTION:Transformer networks were first developed for language transla
 tion and have recently shown great success\, especially with the rise of t
 he GPT family. In this talk\, I will discuss how Transformers can be appli
 ed to data analysis in particle collider experiments. The key idea behind 
 Transformers is the attention mechanism\; I will explain different types o
 f attention\, including self-attention\, cross-attention and sparse attent
 ion. Finally\, I will discuss recent progress in multi-agent AI approaches
  that combine pretrained large language models with particle physics tasks
 . \n\nReferences: arXiv:2401.00452 \, arXiv:2505.03258\, arXiv:2601.21015
  \n\nhttps://indico.ific.uv.es/event/8423/
LOCATION:Universe 1001-Primera-1-1-1 - Paterna. Seminario
URL:https://indico.ific.uv.es/event/8423/
END:VEVENT
END:VCALENDAR