BEGIN:VCALENDAR
VERSION:2.0
PRODID:researchseminars.org
CALSCALE:GREGORIAN
X-WR-CALNAME:researchseminars.org
BEGIN:VEVENT
SUMMARY:Mehrnoosh Sadrzadeh (University College London)
DTSTART:20260522T153000Z
DTEND:20260522T164500Z
DTSTAMP:20260602T185556Z
UID:CompMath/31
DESCRIPTION:Title: <a href="https://researchseminars.org/talk/CompMath/31/
 ">A Quantum Approach to Vision Language Modelling</a>\nby Mehrnoosh Sadrza
 deh (University College London) as part of Relatorium seminar\n\n\nAbstrac
 t\nVision-language models excel at large-scale image-text alignment but of
 ten neglect the compositional structure of language\, leading to failures 
 on tasks that hinge on word order and predicate-argument structure. We sho
 w how techniques from tensor networks and variational quantum circuits hel
 p us solve the problem. To this end\, we  introduce two tools DisCoCLIP an
 d QuCLIP\, multimodal encoders that combine a frozen CLIP vision transform
 er with a tensor network text encoder that explicitly encodes syntactic st
 ructure. We also work with translations of syntax into variational quantum
  circuits. We train both models with a self-supervised contrastive loss an
 d show how the models improve on compositional benchmarks such as SVO-Prob
 es and ARO\, while using a significantly smaller number of parameters. The
  parameter reduction is a known feature of tensor networks and variational
  quantum circuits\, and for this case\, was on average from hundreds of mi
 llions to tens of thousand.\n\nSpeaker's bio: Mehrnoosh is a Professor of 
 CS\,  leads UCL CS's Quantum Learning Labs\, and is the CS Director of Res
 earch.  Her research is supported by a Royal Academy of Engineering (RAEng
 ) Research Chair\, jointly with the BBC and Quantinuum Ltd. Mehrnoosh’s 
 UG and MSc studies were in Sharif University in Iran. Her PhD in Universit
 y of Quebec at Montreal. Previously\, Mehrnoosh had two RAEng  Industrial 
 Fellowships\, in QMUL and UCL\, an EPSRC Career Acceleration Fellowship in
  Oxford\,  an EPSRS PDRF and a Wolfson College Junior Research Fellowship\
 , also at Oxford. \n\n\nModerator's bio: Ted Theodosopoulos is a mathemati
 cian who\, after working for years in academia and industry\, transitioned
  to teaching at the pre-college level sixteen years ago\, the last eight a
 t Nueva\, where he teaches math and economics. Ted’s research background
  is in the area of interacting stochastic systems\, with particular applic
 ations in biology and economics.\n
LOCATION:https://researchseminars.org/talk/CompMath/31/
END:VEVENT
END:VCALENDAR