Benchmarking in AI

On 12 September, researchers and practitioners from across Norway presented talks on benchmarking from different perspectives, including statistical approaches, applications in health care, and the challenges of distributional shift. The program also featured a keynote on safe and secure AI, as well as a tutorial session that will give participants the chance to work hands-on with benchmarking methods.

Photos of the six speakers

You'll find recordings of the session as well as the full program below.

 

Program:

TimeTitlePresenter
08:00 - 08:05IntroductionAndrea Gasparini, NORA.ai
08:05 - 08:45Benchmarking: The key to safe and secure AIMichael Riegler, Simula
08:45 - 08:55Coffee break 
08:55 - 09:35Benchmarking LLMs for NorwegianVladislav Mikhailov, LTG, UiO
09:35 - 09:55AI benchmarking from a statisticians perspectiveHugo Lewi Hammer, OsloMet
09:55 - 10:15AI benchmarking in practice in the Norwegian health care systemHilde Lovett, Helsedir
10:15 - 10:35The limits of bench-marking in the context of distributional shiftBirk Torpmann-Hagen, UiT
10:35 - 10:45Coffee break 
10:45 - 11:30Evaluating AI with AI: LLMs in Benchmarking Pipelines (Tutorial)Sushant Gautam, SimulaMet

 

Published Sep. 25, 2025 2:37 PM - Last modified Oct. 9, 2025 12:57 PM
E-mail this page