Skip to content

8DM60 · Exam skeleton#

Written exam · 28 October13:30–16:30Closed book

Topics, possible questions, and the milestones to reach. Nothing is explained until you ask for it.

Pass conditions#

Component Weight
Multiple-choice questions 20%
Group-specific project questions 15%
General open/problem-based questions 50%
Project poster + code 15%

Written exam ≥ 5.5 · final grade ≥ 6. Lecture 7 is not exam material.

Topics#

  1. Language models — tokens, Transformers, adaptation, evaluation.
  2. Vision models — DINO, CLIP, generative VLMs, validation.
  3. Omics models — scVI, scGPT, CPA, GEARS, perturbation evaluation.
  4. Models for hospital data — patient representations, bias, calibration, generalization.
  5. Clinical workflow integration — deployment, thresholds, validation, governance, economics.
  6. Agentic AI — tools, ReAct, workflows, agents, trajectory evaluation.
  7. Your project — the poster, code, methods, and choices you must defend.

How this grows#

Select Ask beside one question to add only that explanation. Use Ask Claude Code for a UI change. Claude Code edits these Markdown pages directly and rebuilds the site.