BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//wp-events-plugin.com//7.3.7.4//EN
TZID:Asia/Kolkata
X-WR-TIMEZONE:Asia/Kolkata
BEGIN:VEVENT
UID:210@cds.iisc.ac.in
DTSTART;TZID=Asia/Kolkata:20260717T110000
DTEND;TZID=Asia/Kolkata:20260717T120000
DTSTAMP:20260703T060204Z
URL:https://cds.iisc.ac.in/events/seminar-cds-102-july-17th-1100-towards-r
 eliable-context-aware-and-collaborative-multilingual-intelligence/
SUMMARY:{Seminar} @ CDS: #102\, July 17th: 11:00: "Towards Reliable\, Conte
 xt-Aware\, and Collaborative Multilingual Intelligence."
DESCRIPTION:Department of Computational and Data Sciences\nDepartment Semin
 ar\n\n\n\nSpeaker : Dr. Sweta Agrawal\, Senior Research Scientist at Googl
 e DeepMind\nTitle : Towards Reliable\, Context-Aware\, and Collaborative M
 ultilingual Intelligence\nDate &amp\; Time : July 17th\, 2026 (Friday)\, 1
 1:00 AM\nVenue : # 102\, CDS Seminar Hall\n\n\n\nABSTRACT:\nLarge Language
  Models have fundamentally transformed Natural Language Generation\, yet t
 hey remain limited by two major bottlenecks: a static "train-once-and-depl
 oy" learning paradigm and a generic "one-size-fits-all" approach to text g
 eneration. Effective communication is neither static nor generic. It must 
 adapt to specific domains\, evolving knowledge\, and precise audience need
 s across languages. In this talk\, I will present my research on bridging 
 the gap between general-purpose AI capabilities and the nuanced user-centr
 ic needs of real-world applications.\n\nA central theme of my work is enab
 ling multilingual systems to move beyond generic generation by grounding o
 utputs in context. I will highlight my core contributions in controlling a
 udience-specific text complexity\, leveraging context for machine translat
 ion in challenging domains\, and embedding user preferences directly into 
 the generation process\, all under realistic resource constraints. Because
  standard evaluation metrics often fail to capture practical utility in th
 ese settings\, I will also present robust\, user-centric evaluation framew
 orks that measure whether a model's output actually meets the desired cons
 traints and satisfies user needs.\n\nLooking toward the next frontier\, I 
 will outline my goal to transition LLMs from passive text processors into 
 active\, collaborative and intentional agents. This requires moving beyond
  handling explicitly provided contexts to navigating implicit\, open-ended
 \, and diverse multicultural contexts prevalent in real-world scenarios. I
  will discuss how we can build this next generation of reliable language t
 echnologies by equipping models with the capacity to deliberately integrat
 e contextual information across multiple levels — from prompt to interna
 l representation to model parameters — and the modularity to adapt to th
 is evolving knowledge. Transitioning to this paradigm requires fundamental
 ly rethinking how we measure success. I will detail how we can construct d
 ynamic evaluation frameworks that assess long-horizon trajectories\, knowl
 edge retention\, and real-world task success derived from implicit user fe
 edback.\n\nBIOGRAPHY:\n\nDr. Sweta Agrawal is a Senior Research Scientist 
 at Google DeepMind\, where she focuses on scaling and diversifying pretrai
 ning data to ensure Gemini delivers high-quality performance across global
  languages. Before that\, she was a Postdoctoral Researcher at the Institu
 to de Telecomunicações. She earned her Ph.D. in Computer Science from th
 e University of Maryland\, College Park. Her works have received an EMNLP 
 Outstanding Paper Award and top rankings at WMT evaluation campaigns. She 
 has also held research roles at Meta\, Unbabel\, and Adobe\, and actively 
 serves as an Area Chair for leading AI/ML/NLP conferences.\n\nHost Faculty
 : Dr. Danish Pruthi\n\n\n\nALL ARE WELCOME
CATEGORIES:Events,Talks
END:VEVENT
BEGIN:VTIMEZONE
TZID:Asia/Kolkata
X-LIC-LOCATION:Asia/Kolkata
BEGIN:STANDARD
DTSTART:20250717T110000
TZOFFSETFROM:+0530
TZOFFSETTO:+0530
TZNAME:IST
END:STANDARD
END:VTIMEZONE
END:VCALENDAR