Photo: Jesper Eriksson
My research explores how randomness can be used as a resource in computation. Even simple algorithms can display surprisingly rich and powerful behaviour when randomness is built into how they represent and process information. I am especially interested in what this means for artificial intelligence and machine learning. A major part of my research focuses on hyperdimensional computing, a framework that uses high-dimensional random representations for learning, reasoning, and knowledge processing. I study how this framework can be connected to emerging low-power computing hardware, and how it may inspire new types of artificial neural network architectures.
I am interested in unconventional approaches to information processing and machine learning, especially methods that use distributed representations, associative memory and dynamics to compute efficiently.
Interests: Numismatics, phaleristics, freestyle wrestling and reading.
Other: I collected mammoth fossils as a child.
I am particularly interested in issues concerning the attractiveness of the researcher profession and the conditions of research, as well as how we can develop models for long-term scientific excellence. I am also interested in how new technical tools affect society and the role researchers have in this development.
Latest news from the Young Academy of Sweden, published approximately six times a year. You can unsubscribe at any time.
