Cookies on this website

We use cookies to ensure that we give you the best experience on our website. If you click 'Accept all cookies' we'll assume that you are happy to receive all cookies and you won't see this message again. If you click 'Reject all non-essential cookies' only necessary cookies providing core functionality such as security, network management, and accessibility will be enabled. Click 'Find out more' for information on how to change your cookie settings.

Accept all cookies Reject all non-essential cookies Find out more

Collaborators

Colleges

Timo Flesch

DPhil Candidate

  • Human Information Processing (Summerfield Lab)

Representation Learning for Continual Task Performance

Research Interests

I am a DPhil student in Christopher Summerfield's Human Information Processing lab, funded by a Medical Science Division Graduate Studentship (Department of Experimental Psychology and Medical Research Council) and co-supervised by Chris Summerfield  and Andrew Saxe.

A hallmark of human cognition is the ability to learn and execute multiple tasks in succession. The neural basis of this so-called continual learning, however, remains enigmatic. Notably, in Machine Learning, even state-of-the art algorithms fail to continuously acquire structured, protected and reusable representations of the world ("catastrophic interference").

My DPhil research investigates the computational mechanisms and neural representations underlying
continual learning in the human brain. More specifically, I seek to understand  (1) how novel categorisation rules are learned for a fixed stimulus set, without causing interference with previously acquired rules, (2) how abstract representations of rules are reused in novel, but similar contexts (imagine learning the concept of ripeness and applying it to a novel type of fruit) and (3) how unsupervised statistical learning aids as representational scaffold for task learning later in life.

 To reach this goal, I use a combination of behavioural testing, non-invasive neuroimaging and artificial neural network modeling.


Recent publications

More publications

Biography

I hold a BSc in Cognitive Science from Universität Osnabrück in Germany and a PGDip in Computational Statistics and Machine Learning from University College London.

During my time in Osnabrück, I was given the opportunity to work as teaching assistant for undergraduate courses in Mathematics, Logic Programming and Machine Learning. I spent a semester abroad in Chris Summerfield's lab, where I investigated EEG-correlates of perceptual decision making.

Following my undergraduate studies, I moved back to Oxford to work as Research Assistant in Chris Summerfield's lab and conducted research on abstract category learning in humans under the supervision of Chris Summerfield and Hamed Nili. Around that time, I studied towards a postgraduate degree in Computational Statistics and Machine Learning at UCL.

This autumn (2018) I have returned to Oxford to start a DPhil in the Human Information Processing lab.