Winter School
The Winter School was kicked off with two insightful tutorials on multimodal learning by Kristoffer Wickstrøm, Michael Kampffmeyer & Rwiddhi Chakraborty (UiT) and a third tutorial by Samia Touileb (UiB) about ethical challenges in natural language processing (NLP).
Several Winter School participants showcased their research during a dedicated poster session, spanning topics such as AI-enabled vessel segmentation, algorithms for real-time UAV flight trajectory corrections, and fine-tuning large language models for adversial attacks.
The winter school concluded on Friday, with two tutorials on large language models-based agents and co-pilots by Mihaela van der Schaar & Anders Boyd (University of Cambridge & GGD Amsterdam) and a presentation titled Beyond Euclidean Deep Learning by Aiden Durrant (University of Aberdeen).
Main Conference (Tuesday - Thursday)
NLDL - Day 1
After a warm welcome by Prorector for Research Jørgen Berge and General Chairs Robert Jenssen and Michael Kampffmeyer (UiT), NLDL 2026 started with a keynote on self-supervised learning by Professor Yuki Asano (University of Technology Nuremberg).
The session was followed by scientific showcases from several NLDL 2026 participants, in the form of oral and poster sessions.
In the afternoon, a side event on diversity in AI took place, which delved deeper into how AI inadvertently can reflect and perpetuate societal biases, and how misinformation or flawed data can lead to biased AI systems. This event featured the following experts: Alex Moltzau, Team Lead at the Norwegian Government AI Unit, Norwegian Digitalisation Agency (Digdir), Eva Breznik, Postdoctoral Researcher at KTH Royal Institute of Technology, Artem G., Researcher at UiT and Integreat - Norwegian Centre for Knowledge-driven Machine Learning and Elisabeth Wetzer, Associate Professor at UiT and SFI Visual Intelligence (Moderator).
The day concluded with a mingling session and a trip to the idyllic Sommarøy Island outside of Tromsø; with a recap of a 100 years of northern lights research by Researcher Andreas Kvammen, followed by tapas and good conversations.
NLDL - Day 2
The second day started with a keynote by Devis Tuia, Associate Professor at EPFL: Masked Autoencoding Enables Multimodal Learning for Environmental Sciences.
Tuia presented recent advances in environmental data analysis based on the integration of multiple data streams, synergized thanks to masked autoencoders.
After the first keynote lecture, a poster session featured newly published AI research on topics such as foundation models for medical images, uncertainty in 3D segmentation, and spatio-temporal landmark detection.
Later on, Professor Ira Assent at Aarhus University gave a keynote lecture about robust explanations for classification, clusters, and outliers.
Assent highlighted recent work in explainable AI on robust and trustworthy explanation methods that handle complex learned decision boundaries.
After a stimulating second day at NLDL 2026, participants gathered for the conference dinner at Scandic Ishavshotel, which included speeches from Program Chair Benjamin Ricaud and General Chair Robert Jenssen, who thanked the participants for travelling to Tromsø to showcase the latest AI research advances in AI and deep learning.
NLDL - Day 3
The third and last day at NLDL 2026 featured a high quality series of oral and poster sessions, the final keynote, and two fascinating AI panel discussions:
The first keynote by Shaheen Syed (Principal Data Scientist at Equinor) was about the industrial AI implementation at Equinor. Syed's keynote examined actual implementations of deep learning models in industrial settings, showing how these technologies solve practical problems at scale.
The talk was followed by the third and final poster session – featuring recently published AI research on topics such as AI prediction of calving events in Antarctica, large language models for low-resource languages, and oil spill detection via uncertainty-aware deep learning
The day finished with NORA's Panel Discussion on the European AI in Science Strategy – a fruitful discussion between panelists Georgios L., Aasa Feragen, Robert Jenssen and moderator Michael Solvang.The NORA discussion delved deeper into these experts' bottom-up perspectives on how the strategy's top-down vision reflects the realities of AI-driven science today.
Industry Events
Industry & AI was hosted on Thursday in the afternoon, and addressed important subjects such as why artificial intelligence is important for industry, how it can provide additional value to industry, problems that AI can potentially solve, and how academia and industry should collaborate on facilitating the safe and sustainable use of AI.
The event consisted of four short keynotes given by the experts below in the field, and ended with a panel discussion. The keynotes were:
- Cathrine Tegnander, Digital Subsurface Manager at Aker BP ASA
- Erland Grimstad, Co-Founder at Abdera.ai
- Njål Borch, Senior AI Engineer at Schibsted
- Bjørn Fjukstad, Product Manager at DIPS AS (panel moderator)
In the evening, a second industry event took place at at Storgata Camping - a cozy pub in Tromsø. It was the second edition of Meet the AI Industry, an NLDL 2026 side event hosted by SFI Visual Intelligence and UiT The Arctic University of Norway in collaboration with the Norwegian Artificial Intelligence Research Consortium (NORA).
This event aimed to provide a broader view of where in the industry AI skillsets are required. Often one might only hear about larger companies such as Microsoft, Amazon, etc, but so much more of our industry sector needs AI expertise.
Each industry partner gave a small presentation of their company, what they do and where they exist in the world. After the presentations there was plenty of time for networking.
Speakers at the event were:
- Johan Söderberg (MathWorks)
- Lars Adrian Giske (iTromsø)
- Astrid Pettersen (AepTech)
- Njål Borch (Schibsted)
- Johannes Nilsen (Abdera)
- Bjørn Fjukstad (DIPS)
- Ante Grgurević (Visage Technologies)