Become a Mentor
Why mentor?
Make progress on your own research questions that you otherwise wouldn’t have capacity to
Gain unique expertise on your project, as your mentees may have a background you lack
Help ambitious Oxford students gain impactful careers
What’s the background of the mentees?
Last round we received 63 applications for 16 places so can afford to be selective based on your preferences.
However, typically selected candidates were late undergraduate or masters students, mostly in maths or computer science.
How do you select mentees?
We check the background and motivation of candidates through an online application against the criteria you specify. You are welcome to add a longer question specific to your project to check their interest, background and thoughtfulness about the project.
We interview shortlisted candidates to cross check their answers and assess their motivation.
We will send you a shortlist of candidates for you to select from. You are welcome to take any number of shortlisted candidates (including 0).
What do we ask from you?
Write a ~500 word project brief (see the old project briefs for example). This can be any self contained project that you think can be impactful.
Attend our in person kick off day in Oxford (travel expenses paid).
Roughly 2 hours a week for 8 weeks, comprising roughly 1 hour of meetings and 1 hour of reading student’s updates.
When is it?
Our in person kick off day on Sunday 25th October 2026, 1-6pm with travel and food paid for
Then the programme runs for 8 weeks, starting Monday 26th October (although mentors have often worked with their teams for longer.
What does the mentor application process look like?
Application (requiring brief outline of project)- due 20th September
Interview w/c 21st September to ensure it’s a good fit on both sides
Send us your final 500 word project proposal
Offer
What do our past projects look like?
Winning paper last round: AI-Powered Cyber Attacks and AI Consciousness: Incident Scenarios and Potential Mitigations
Winning paper in the 1st round: Training Invariance-Aware Diverse Reward Model Ensembles using STARC