The team primarily focuses on participation in public machine learning challenges.
The projects the team will work on are determined in mutual agreement.
Some potential target challenges include:
The ARC prize: The 'abstraction and reasoning corpus' (ARC) is a well-known public benchmark that intends to push the boundaries of AI on abstract reasoning (2024, 2025, 2026).
The AIMO prize: The 'AI for Mathematical Olympiad' (AIMO) prize is another highly-regarded public challenge that aims to develop machine learning models that can solve mathematical problems (for the Mathematical Olympiad). (2024, 2025, 2026)
Machine learning conference competitions: Several machine learning conferences host yearly competitions. For example, NeurIPS, one of the biggest conferences in machine learning, host yearly challenges (2023, 2024, 2025). The European Conference on Machine Learning also holds yearly 'Discovery Challenges' (2024, 2025).
AI for (bio)science: There is a growing interest in 'AI for science', i.e., the use of AI to achieve progress in other scientific fields (main example). Given our position at the Leiden Bioscience Park, there should be opportunties to connect to researchers from other disciplines, such as chemistry and biology. Several recent ML challenges focus in this direction, for example on predicting properties of polymers, the 3D-structure of RNA, the responses of biological cells or the genes involved in obesity.
The (Dutch) AI Cup: A regional competition initiated by the Delft AI Student Team 'Epoch' (2026).
AI Hackathons: For a hackathon you actually physically go to a location with a team to work on a challenge for a couple of days. One example is the recurring 'Hackathon for Good' (in The Hague).
'AI for Good' community: Organized from TU Eindhoven, there is a community around challenges with positive societal impact, called FruitPunch.ai (announcement).