Integreat PhD Fellow - UiT The Arctic University of Norway, Department of Physics and Technology
Academic interests
Daniel’s research concerns latent spaces in deep neural networks and in particular LLMs. He investigates how we can understand and modulate the high dimensional information they encode through computational methods. The goal of his effort is to attain better ways to control how models behave.
Background
Previously Daniel researched the novel application of Machine Learning methods to financial statements at the Oxford-Man Institute for Quantitative Finance within an academic focus on Deep Learning since 2016. Following the COVID pandemic, he took a break in industry to work on applied machine learning problems in areas such as biometric security, news media, private banking, and CNC machining.
PhD project
"Exploration and control of the inner representation in generative AI models"
Supervisors
- Arnoldo Frigessi, University of Oslo, Department of Mathematics
- Benjamin Ricaud, UiT The Arctic University of Norway, Department of Physics and Technology
- Ali Ramezani-Kebrya, University of Oslo, Department of Informatics
Related articles
How smart are chatbots, actually? A new method could give us the answer