The PhD defence and trial lecture will be streamed. The chair of the defence will moderate the disputation.
Ex auditorio questions: the chair of the defence will invite the audience physically present in the auditorium to ask ex auditorio questions.
→ Live streaming of trial lecture and disputation
→ Request for thesis copy (available until the disputation starts)
Trial lecture
Time and place: October 30, 2026 - 08:00 AM
Lille Fysiske auditorium (V232) - Fysikkbygningen
Title: "Energy efficiency of different neural models"
Main research findings
Can general intelligence be created, not from giant AI models, but from many small units interacting? Our brains may look like single entities, but they are actually intricate hierarchies of interacting parts. From molecules to cells, from cells to organs, small parts interact locally and create complex behavior at the level above. This is called collective intelligence. In this thesis, I look at just such a system in the brain: Repeated cell-structures that process our senses, called cortical columns. I create AI systems inspired by cortical columns, building on the architecture of Neural Cellular Automata (NCAs), which essentially consists of many parallel AI modules working together. When these modules learn together, without assigned roles, they develop a general function: Thus, I dub them generalist modules. In teaching the NCA to do similar tasks to the cortical columns, such as sensing objects, I find that the system learns emergent attention to important parts of the object. I also find that such systems in general are scalable, robust, small, and cost-effective. Finally, I present some analysis of these systems, working towards more explainable AI. The goal is to understand the properties of such repeated brain modules, for greener and more general AI.
Adjudication Committee
- Professor Gunnar Tufte, Department of Computer Science, Norwegian University of Science and Technology (NTNU); Norway
- Associate Professor Lana Sinapayen, National Institute for Basic Biology, Sony Computer Science Laboratories Kyoto, Japan
- Professor Lasse Clausen, Department of Physics, University of Oslo, Norway
Supervisors
- Associate professor Mikkel Lepperød, Department of Physics, University of Oslo, Norway
- Assistant Professor Sidney Pontes-Filho, Oslo Metropolitan University, Norway
- Associate Professor Kai Olav Ellefsen, Department of Informatics, University of Oslo, Norway
- Professor Kyrre Harald Glette, Department of Informatics, University of Oslo, Norway
- Professor Anders Malthe-Sørenssen, Department of Physics, University of Oslo, Norway
Chair of defence
Head of Department Susanne Friederike Viefers, Department of Physics, University of Oslo, Norway
Candidate contact information
LinkedIn: Mia-Katrin Kvalsund
E-mail: mia.kvalsund@gmail.com
Contact information to Department: Line Trosterud Resvold