AI Safety Amsterdam

Research Hub

We connect students and researchers to work on meaningful AI safety problems. Whether you want to join an active research group, find supervision for your own project, or collaborate with others in the community — this is where that starts.


Get Involved

I'm looking for a supervisor

Join an existing research group or bring your own project idea. We'll help you find a supervisor and connect you with others working on similar problems.

Contact us

I'm looking to supervise

PhD students, professors, and industry researchers are welcome. If you want to supervise a group or propose a project for students, we can handle coordination and help you find motivated people.

Contact us

Meet our Supervisors

Our supervisors are researchers and professionals who guide student groups and propose projects.

Leon Eshuijs

Leon Eshuijs

Supervisor

Mechanistic Interpretability

ELLISMech Interp
Research Agenda

Research Outputs

Selected work produced by AISA members through research groups, hackathons, and independent projects.

  • MechInterp Research (Ongoing)

    Research into mechanistic interpretability of neural networks, exploring how models represent concepts internally and how that can inform safety.

    MechInterp In progress

This section will grow as our research matures. If you're an AISA member with work to feature, get in touch.