BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Memento EPFL//
BEGIN:VEVENT
SUMMARY:IC Colloquium: Towards data-driven modeling in large-scale natural
 istic neuroscience
DTSTART:20230328T120000
DTEND:20230328T130000
DTSTAMP:20260407T061544Z
UID:ad9a04992c98bae95cca05deac48c9d93bd99366ae8fb0531651f430
CATEGORIES:Conferences - Seminars
DESCRIPTION:By: Meenakshi Khosla - MIT\nIC/SV Faculty candidate\n\nAbstrac
 t\nNeuroscience is currently undergoing an explosion in the availability o
 f large-scale brain activity data\, so the major challenge no longer lies 
 in data collection\, but in deriving understanding from this abundant stre
 am of complex\, high-dimensional\, noisy data with methods that fully leve
 rage its potential. How can we understand neural representations and infer
  computational principles from large-scale brain activity data directly? M
 y research group will develop the theory\, modeling\, and machine learning
  techniques needed to tackle these challenges\, and will provide algorithm
 ically and computationally precise accounts of information processing in t
 he brain. Progress in this research could advance our understanding of bio
 logical intelligence and the neural basis of neuropsychiatric disorders\, 
 inform novel treatments and neural prostheses\, and lead to the developmen
 t of new approaches to generating machine intelligence.\n \nIn this talk\
 , I will present lines of previous and proposed research that highlight th
 e potential of this vision. First\, I will present a line of data-driven m
 odeling that revealed the representational structure in the high-level vis
 ual cortex and led to the discovery of a neural population selectively res
 ponsive to images of food. Second\, I will present a modeling framework\, 
 called response-optimization\, for inferring computations directly from br
 ain activity data with minimal apriori hypotheses. Here\, we trained artif
 icial neural network (ANN) models directly to predict the brain activity r
 elated to viewing natural images. We then developed techniques for interpr
 eting the networks and characterizing the emergent functional capabilities
  of these brain response-optimized networks. This work highlights how mode
 ls trained to capture human brain activity can spontaneously recapitulate 
 human-like behavior. Third\, I will present my work on developing neural n
 etwork models of brain responses across wide-spread cortical regions to dy
 namic\, multi-modal stimuli like movies\, with an integrated modeling appr
 oach that captured visual attention\, multi-sensory auditory-visual intera
 ctions and temporal context. Finally\, I will propose future directions fo
 r leveraging these data-driven computational tools toward i) understanding
  the representational structure underlying our capacity for high-level cog
 nitive processes like social perception\, ii) understanding the granularit
 y of similarity between biological and artificial neural networks\, iii) u
 nderstanding the functional consequences of brain alignment in ANN represe
 ntations\, and iv) understanding not just how the brain is organized or ho
 w a process works but fundamentally\, why the brain works the way it doe
 s.\n\nBio\nMeenakshi Khosla is a Postdoctoral associate in the Brain and C
 ognitive Sciences department at MIT. Her research interests lie at the int
 ersection of neuroscience\, artificial intelligence\, and large-scale data
  analysis. Her current projects focus on leveraging large-scale data to de
 velop interpretable machine learning tools for understanding structured ne
 ural representations and computations in biological systems.  Previously
 \, Meenakshi received her PhD in Electrical and Computer Engineering (ECE)
  from Cornell University\, where she worked broadly at the intersection of
  machine learning and neuroimaging\, developing predictive models to under
 stand the distinctive characteristics of the brains of people affected wit
 h different mental disorders. She completed her undergraduate degree in e
 lectrical engineering at the Indian Institute of Technology Kanpur.\n\nMor
 e information
LOCATION:https://epfl.zoom.us/j/67652538478
STATUS:CONFIRMED
END:VEVENT
END:VCALENDAR