Jelena Sucevic

Jelena Sucevic

Postdoctoral Research Scientist

University of Oxford

Hi there, welcome!

My research explores how learning occurs in human minds and in machines. I want to understand the algorithms and building blocks of learning. I study how fundamental cognitive abilities such as attention, memory, and language make learning, and ultimately intelligence, possible.

I am particularly interested in the process of learning to learn, and how the developing brain masters complex cognitive functions. I also use biologically-inspired neural networks to uncover the learning mechanisms in the brain.

In addition, I’m curious about how principles of early learning can be used to inspire novel machine learning algorithms, especially in the domain of multimodal and weakly supervised learning. As Turing pointed out, “instead of trying to produce a programme to simulate the adult mind, why not rather try to produce one which simulates the child’s?"

Interests
  • Learning
  • Cognitive Development
  • Neural network modelling
  • Cognitively-inspired AI
Education
  • PhD in Experimental Psychology, 2018

    University of Oxford

  • Visiting Researcher, 2018

    Harvard Medical School

  • Visiting Student, 2016

    University of Zurich

  • MSc in Psychological Research, 2013

    University of Belgrade

  • Visiting Student, 2013

    University of Goettingen

  • BSc in Psychology, 2012

    University of Belgrade

Ongoing Projects

Multimodal learning: language as a learning signal during information selection and integration
Learning words in a structured world: the interface between statistical learning and word mapping
A neural network model of category learning

Recent Publications

(2022). A neural network model of hippocampal contributions to category learning. biorXiv preprint.

(2022). Discovering category boundaries: The role of comparison in infants’ novel category learning. Infancy.

(2021). Can algorithms learn from babies? Exploring how infant learning can inform and inspire unsupervised learning algorithms. Proceedings of the Annual Meeting of the Cognitive Science Society..

(2021). The role of labels and motions in infant category learning. Journal of Experimental Child Psychology.

(2020). What can babies teach us about contrastive methods?. NeurIPS.

(2020). Let’s talk action: Infant-directed speech facilitates infants’ action learning. Developmental Psychology.

Contact