Watch the full Q&A Meeting Recording

Local Site Logistics & Registration

Q: Is there an expected number of participants or teams per local site?
A: The number of participants is entirely up to the local organizer and the room's capacity. Previous sites have hosted anywhere from 5 to over 30 participants depending on how broadly the event was advertised locally. If a physical room reaches capacity, the central registration system can cap sign-ups for that specific local site.
Q: Are there registration fees for participants?
A: No, participation in the event is completely free of charge for everyone. All hackathon activities are designed to be cost-neutral to ensure broad accessibility.
Q: Can existing grants cover local costs, or can we invite local sponsors?
A: Local organizers can use existing grants to cover food and coffee, provided they clear this with their local research office. Local sites are also highly encouraged to invite local sponsors for support or to fund local prizes. We are delighted to highlight any sponsor acknowledgments from local site organizers on the central hackathon website.

Participants & Team Formation

Q: Who can participate, and is machine learning expertise required?
A: Participation is open to everyone—from high school students to senior scientists—provided local institutional rules regarding external visitors and minors are strictly followed. Participants do not need to be machine learning experts. Understanding the domain problem is far more critical, and the event is specifically designed to pair domain experts with coders.
Q: How does team formation work?
A: Participants will gain access to a Miro board the day before the event to self-organize around specific problems. Teams frequently form heterogeneously across global time zones, though pre-formed local teams are also completely acceptable. Teams generally select problems without any noticeable correlation to their geographic site.

Problems, Datasets & Resources

Q: What types of problems and datasets are acceptable?
A: Datasets should ideally focus on microscopy (including optical, mass spectrometry, or scanning probe), but computational models like digital twins are also welcome. Data from previous publications can be used as long as the hackathon workflow and results remain open-source. Very large datasets should be decimated to a representative size suitable for hosting via platforms like Google Colab.
Q: When is the deadline for submitting problems?
A: Organizers ask that datasets and problems be submitted anytime from now until two weeks before the hackathon. To ensure fairness across the competition, all problems are kept confidential and revealed to participants exactly one day before the event begins.
Q: Are computational resources provided, or do local sites need to supply them?
A: Local sites are not expected to provide any special computational resources. Participants generally rely on standard cloud access like Google Colab. The central organizing team is also exploring sponsorships to potentially provide additional cloud computing resources to participants.
Q: How are projects evaluated, and can local organizers join the jury?
A: Submissions are judged by a multi-person panel using a non-disclosed rubric focused on domain understanding, physical significance, and solution robustness. Local site organizers are strongly encouraged to join the jury. Expanding the jury's domain expertise is critical for accurately evaluating the physical significance of the proposed solutions.