Arvutiteaduse instituut
  1. Kursused
  2. 2023/24 sügis
  3. AI-Safety (MTAT.03.317)
EN
Logi sisse

AI-Safety 2023/24 sügis

Older Datamining Seminars: 2008k » 2008s » 2009k » 2009s » 2010k » 2011k » 2012s » 2014k » 2014s » 2014k

  • About
  • Points
  • Project
  • Additional materials
  • Timetable
  • Homeworks
  • Project ideas

Project

The last four weeks of the course is dedicated to working on a small project.

The Alignment course web page proposes many project topics one can choose from.

The project page on the Alignment course page is currently not available but you can look the following materials for inspiration:

  • A somewhat chaotic list of exercises and projects
  • The page that the AI Safety Fundamentals people suggested to consult while the project page is down.

To complete the project part of the course is the following:

  • Choose a project topic that you find interesting or that is somehow useful for your main research. The project cannot be too large, because the expected optimal work load would be 3-4 hours of work per week + seminar attendance.
  • Work on this project on a weekly basis, keeping a research journal of your progress.
  • Attend the seminar to shortly present your progress.
  • Comment and give feedback to others on their progress.

The weekly seminar points will be given for the following tasks/deliverables:

  • 3 points: progress reported in the research journal (can be simply a shared Google doc)
  • 2 points: shortly presenting your progress in the seminar
  • 2 points: in the seminar giving feedback to others

Project tasks: week 0

  • Start your research diary by doing the two exercises given in the materials
  • Big picture plan for your self: if and how you want to be involved with AI safety in your career.
  • Come up with a topic for your project and try to formulate a 4-week plan for completing it using SMART goals.

Project tasks: week 1

  • Work on your project according to the plan. Update the plan if necessary.
  • Record the results and progress of the week's work in your research diary (approximately 1 page, min 0.5, max 2 pages).
  • Arvutiteaduse instituut
  • Loodus- ja täppisteaduste valdkond
  • Tartu Ülikool
Tehniliste probleemide või küsimuste korral kirjuta:

Kursuse sisu ja korralduslike küsimustega pöörduge kursuse korraldajate poole.
Õppematerjalide varalised autoriõigused kuuluvad Tartu Ülikoolile. Õppematerjalide kasutamine on lubatud autoriõiguse seaduses ettenähtud teose vaba kasutamise eesmärkidel ja tingimustel. Õppematerjalide kasutamisel on kasutaja kohustatud viitama õppematerjalide autorile.
Õppematerjalide kasutamine muudel eesmärkidel on lubatud ainult Tartu Ülikooli eelneval kirjalikul nõusolekul.
Courses’i keskkonna kasutustingimused