BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Cultural Biases in LLMs: From Behavioural analysis to Internal Rep
 resentations - Siddhesh Pawar (University of Copenhagen)
DTSTART:20260123T110000Z
DTEND:20260123T120000Z
UID:TALK243340@talks.cam.ac.uk
CONTACT:Lucas Resck
DESCRIPTION:Large language models are increasingly deployed across diverse
  cultural contexts\, yet they often encode and perpetuate subtle cultural 
 biases. In this talk\, I will explore how these biases manifest in practic
 e\, drawing on my work examining name-based cultural stereotyping and the 
 challenges that arise when models encounter users with complex\, multicult
 ural identities. Moving beyond behavioural analysis\, I will then discuss 
 mechanistic approaches to understanding how cultural representations and b
 iases are encoded within model internals\, and how such analyses can infor
 m more targeted mitigation strategies. Finally\, I will present ongoing wo
 rk on understanding how value-based post-training shapes model behaviour a
 cross cultures.\n\nBio: Siddhesh is a PhD student at the University of Cop
 enhagen and currently visiting researcher at the Centre for Human-Inspired
  AI (CHIA) and the Language Technology Lab at the University of Cambridge 
 through the ELLIS fellowship. His research interests lie at the intersecti
 on of cultural NLP\, computational social science\, and interpretability.
LOCATION:GR04 (English Faculty Building\, 9 West Road\, Sidgwick Site) and
  online (https://teams.microsoft.com/meet/3797179294965?p=naCTdc5XVRYf12vL
 ac)
END:VEVENT
END:VCALENDAR