What we do
Learning materials for students, teachers and mentors. Using experiment cards, we explain hypotheses, data origin, fair comparisons and negative results. The course is still being prepared.
Time and a first task
Plan for about four hours a week. A first 30-minute step: design an exercise in which a student looks for missing reproducibility details in a made-up configuration. Use synthetic values only and do not invent model results.
Issues that need help. If none fits, propose an exercise in a new issue.
Make it reusable
- state the required knowledge, time and resources, and what can be done without a GPU;
- keep observations, hypotheses and unfinished runs apart;
- do not require access to restricted datasets.
Material with medical terms is checked by a specialist. See education and the experiment log.