AI Seminar: XAI for representation learning
Join us for a talk by Robert Jenssen, Director of SFI Visual Intelligence and Professor at UiT The Arctic University of Norway. Everybody is welcome to attend.
Title
XAI for representation learning
Abstract
This talk will discuss motivation for recent explainability research in deep learning, so-called XAI. The talk will briefly outline a taxonomy of present XAI methods. For a large part, XAI try to identify input features that drives supervised class predictions and have received widespread attention. These methods are mostly what we refer to as post-hoc, i.e “after-the-fact” explainability. The talk then touches upon some recent alternatives aimed to be self-explainable, i.e methods with a built-in capability for providing explanations. The main aim of the talk is however to introduce the first XAI work published for explainability in representation learning. Representation learning, e.g. by self-supervision, is crucial in a wide array of recent applications.
Anyone can participate in the event, but please use this form to register.
This seminar is a part of the AI Seminar Series organised by SCIENCE AI Centre. The series highlights advances and challenges in research within Machine Learning, Data Science, and AI. Like the AI Centre itself, the seminar series has a broad scope, covering both new methodological contributions, ground-breaking applications, and impacts on society.