BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Memento EPFL//
BEGIN:VEVENT
SUMMARY:The role of Instruction-tuned Models as Annotators: Exploring Labe
 l Variation
DTSTART:20231031T110000
DTEND:20231031T120000
DTSTAMP:20260917T070233Z
UID:97602ebdab00279cb99808290c59e2700bf29f1422172a9951004829
CATEGORIES:Conferences - Seminars
DESCRIPTION:Flor Miriam Plaza-del-Arco is a Postdoctoral Research Fellow a
 t the MilaNLP group at Bocconi University (Italy).\nHer research interests
  mainly focus on Natural Language Processing\, particularly in hate speech
  detection\, emotion analysis\, early web risk prediction\, and large lang
 uage models evaluation.\nThe zero-shot learning capabilities of large lang
 uage models (LLMs) make them ideal for text classification without annotat
 ion or supervised training. \n\nany studies have shown impressive results
  across multiple tasks. While tasks\, data\, and results differ widely\, t
 heir similarities to human annotation can aid us in tackling new tasks wit
 h minimal expenses. The ultimate promise of LLMs is that their language ca
 pability lets them generalize to any text classification task. What if the
  answer is not to wait for one model to rule them all\, but to treat their
  variation similar to the disagreement among human annotators? Not as indi
 vidual flaws\, but as specializations we can exploit.\n\nIn this talk\, we
  will explore the potential of state-of-the-art instruction-tuned models t
 o serve as "annotators" across various established NLP tasks\, including s
 entiment classification\, age and gender prediction\, topic classification
 \, and hate speech detection. We will answer two fundamental questions: Do
  we continue to require human annotators\, and do the variations in human 
 labeling also manifest in LLMs? Additionally\, we will discuss the tradeof
 fs between speed\, accuracy\, cost\, and bias when it comes to aggregated 
 model labeling versus human annotation
LOCATION:BC 04 https://plan.epfl.ch/?room==BC%2004 https://epfl.zoom.us/j/
 69499602273?pwd=WTBWK1o0L1Z0b085anBiM094STFjQT09
STATUS:CONFIRMED
END:VEVENT
END:VCALENDAR