BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Iowa State University CALS LAS Web Team//sites.iastate.edu//EN
BEGIN:VEVENT
UID:20260507T180000-4908-www.cs.iastate.edu
DTSTART:20260507T180000Z
SEQUENCE:0
TRANSP:OPAQUE
DTEND:20260507T190000Z
LOCATION:Atanasoff 223
SUMMARY:PhD Preliminary Oral Exam: Shaila Sharmin
CLASS:PUBLIC
DESCRIPTION:From Detection to Mitigation: Numerical Instability Analysis fo
 r Machine Learning SystemsMachine learning systems are increasingly deploy
 ed in safety-critical and resource-constrained environments\, yet their nu
 merical reliability remains poorly understood. This thesis presents a syst
 ematic study of numerical instability in machine learning systems — from
  training programs to quantized neural networks to large language models 
 — addressing three interconnected problems: detection\, analysis\, and r
 epair.\n\nMore information at: https://www.cs.iastate.edu/event/2026/phd-p
 reliminary-oral-exam-shaila-sharmin
DTSTAMP:20260510T173326Z
END:VEVENT
END:VCALENDAR