BEGIN:VCALENDAR
VERSION:2.0
PRODID:https://evenements.uqam.ca
X-PUBLISHED-TTL:P1W
BEGIN:VEVENT
UID:34627@https://evenements.uqam.ca
DTSTART:20261112T103000Z
SEQUENCE:4
TRANSP:OPAQUE
URL:https://evenements.uqam.ca/evenements/seminaire-au-dic-secret-in-the-tr
 aining-data-how-infant-behavior-solves-learning-problems-par-linda-b-smith
 /34627?date=2026-11-12_10-30-00
LOCATION:UQAM - Pavillon Président-Kennedy (PK) (201\, avenue du Présiden
 t-Kennedy\, Montréal )
SUMMARY:Séminaire au DIC: «Secret in the training data: How infant behavi
 or solves learning problems» par Linda B. Smith
CLASS:PUBLIC
DESCRIPTION:Séminaire ayant lieu dans le cadre du Doctorat en informatique
  cognitive\, en collaboration avec le centre de recherche CRIA      
         \n\n\n \n\n\nTITRE : Secret in the training data: How in
 fant behavior solves learning problems\n\n\n \n\n\nLinda B. SMITH\n\n\nJe
 udi le 12 novembre 2026 à 10h30\n\n\nLocal PK-5115 (Il est possible d'y 
 assister en virtuel en vous inscrivant ici)               
   \n\n\n \n\n\nRÉSUMÉ\n\n\nMuch of the information in the world is la
 tent: it is not revealed without action by the perceiver. This creates dir
 ect\, real-time relations between sensory input and behavior\, between the
  current and next states of the learning system\, and between inputs and t
 he outcomes of actions on the world. I will present findings from our anal
 yses of the visual statistics of infants’ egocentric images\, collected 
 at the scale of daily life in the home. These findings suggest that the qu
 ality and spatiotemporal structure of infants’ self-generated input cont
 ribute to their efficient learning about visual objects. When human learni
 ng is more efficient than current learning mechanisms can explain\, theori
 sts often posit intrinsic “inductive biases” that constrain learning o
 utcomes\, enabling faster and more reliable learning from complex\, variab
 le\, and noisy training data. The visual statistics generated by infants a
 nd toddlers interacting with their everyday world reveal another source of
  constraint: the learner’s own behavior structures and biases the traini
 ng data itself\, rather than merely constraining the inferences drawn from
  it. These findings may suggest principles for designing training data tha
 t support efficient learning even in machines whose learning mechanisms di
 ffer from those of humans.\n\n\n \n\n\nBIOGRAPHIE\n\n\nLinda B. SMITH is 
 Distinguished Professor of Psychological and Brain Sciences at Indiana Uni
 versity – Bloomington\, IN.Taking a complex systems perspective\, she se
 eks to understand the interdependencies among perceptual\, motor and cogni
 tive developments during the first three years of post-natal life.\,  She
  uses wearable sensors to study how the young learner’s own behavior cre
 ates the statistical structure of the learning environments  at the scale
  of everyday life AI.\n\n\n \n\n\nRÉFÉRENCES \n\n\nRamirez\, F. M.\, C
 lerkin\, E. M.\, Crandall\, D. J.\, &amp\; Smith\, L. B. (2026). A solutio
 n to generalized learning from small training sets found in infants’ rep
 eated visual experiences of individual objects. Proceedings of the Nation
 al Academy of Sciences\, 123(33)\, e2536143123. https://arxiv.org/pdf/251
 0.15060\n\n\nPetroff\, Z. J.\, Jayaraman\, S.\, Smith\, L. B.\, Candy\, T.
  R.\, &amp\; Bonnen\, K. (2025). The world through infant eyes: Evidence f
 or the early emergence of the cardinal orientation bias. PNAS\, 122(16)\,
  Article e2421277122. https://doi.org/10.1073/pnas.2421277122.  \n\n\n
 Sheybani\, S.\, Hansaria\, H.\, Wood\, J. N.\, Smith\, L. B.\, &amp\; Tiga
 nj\, Z. (2023). Curriculum Learning with Infant Egocentric Videos. Paper 
 to be presented at NeurIPS 2023. \n\n\nKarmazyn-Raz\, H. &amp\; Smith\, L
 .B. (2023) Sampling statistics are like story creation: A network analysi
 s of parent-toddler exploratory play. Philosophical Transactions of the R
 oyal Society B \n\n\nSmith\, L. B.\, Jayaraman\, S.\, Clerkin\, E. &amp\;
  Yu\, C. (2018). The Developing Infant Creates a Curriculum for Statistica
 l Learning.Trends in Cognitive Sciences\, 4\, 325-336. PMID: 29519675 PMCI
 D: PMC5866780 \n\nMot-clés : Séminaire\, LLMs\, NeuroQAM\, Neurolab\, L
 ATECE UQAM INFORMATIQUE\, LATECE\, CRIA\, Département de Neuroscience\, P
 hilosophie\, Sciences cognitive\, École de langues\, Sciences cognitives\
 , neurosciences cognitives\, Institut des sciences cognitives\, Apprentiss
 age du langage naturel\, sciences du langage\, apprentissage machine\, app
 rentissage profond\, langage automatique\, langage cognitif\, metacognitio
 n\, Cognition humaine\, Cognition\, intelligence artificielle\, IA\; intel
 ligence artificielle\; société\, IA \; intelligence artificielle\, soci
 été\, IA\, intelligence artificielle\, chatGPT\, enseignement supérieur
 \, IA\, intelligence artificielle\, département de linguistique\, départ
 ement de psychologie\, doctorat en psychologie\, Maîtrise en informatique
 \, Faculté des sciences humaines\, Faculté des sciences de l'UQAM\, Facu
 lté des sciences\, Département d'informatique\, doctorat en informatique
  cognitive\, doctorat en informatique\n\nPrix : Gratuit\n\n
CATEGORIES:Séminaire,Conférence
DTSTAMP:20261023T082620Z
CREATED:20261013T185238Z
LAST-MODIFIED:20261014T141741Z
END:VEVENT
END:VCALENDAR