Software Project: Implementing Mechanistic Interpretability in NLP
Software Project, Saarland University, Winter Semester 2026/27, 2026
Wednesday 16:15 to 17:45
Kick-off: 21.10.2026
End: 31.03.2027
ECTS: 8 for M.Sc. students, 12 for B.Sc. students
B.Sc. students will only be admitted to the course with sufficient prior expertise and English skills.
Course Content
This course will consist of two parts:
In the first part, we will focus on understanding and using the most important methods of mechanistic interpretability. Every week will be dedicated to a method, with a mini-lecture, a hands-on part, and a small homework.
In the second part, students will work on a project in small groups to get hands-on experience in conducting independent research in mechanistic interpretability. There will be weekly office hours and a final presentation. At the end of the semester, students submit a short report and a GitHub repo.
Preliminary Schedule
| Date | Content |
|---|---|
| 21.10.2026 | Kick-off and Introductory Lecture on Mechanistic Interpretability |
| 28.10.2026 | No Meeting |
| 03.11.2026 | Alternative Meeting Slot, Tuesday, 16:15, Topic 1 - tbd |
| 11.11.2026 | Topic 2 - tbd |
| 18.11.2026 | Topic 3 - tbd |
| 25.11.2026 | Topic 4 - tbd |
| 02.12.2026 | Topic 5 - tbd |
| 09.12.2026 | Topic 6 - tbd |
| 16.12.2026 | Topic 7 - tbd |
| 23.12.2026 | No Meeting |
| 30.12.2026 | No Meeting |
| 06.01.2027 | Final Project Development |
| 13.01.2027 | Final Project Development |
| 20.01.2027 | Office Hour |
| 27.01.2027 | Office Hour |
| 03.02.2027 | Office Hour |
| 10.02.2027 | Office Hour |
| 17.02.2027 | Office Hour |
| 24.02.2027 | Office Hour |
| 03.03.2027 | Final Presentations |
| 31.03.2027 | Deadline: Report and Code |
A well-meaning tip: Please expect this course to be about as much work as a B.Sc. Thesis (compare to ECTS). Speaking from experience of the last few iterations of this course, doing a software project half-heartedly is not a fun experience. If this sounds like too much work because of other responsibilities, it probably is. However, also speaking from experience, motivated students can learn a LOT in this course, and some even came out with publications. I personally do not care about the level of experience you come in with - as long as you are willing to put in the work and are eager to learn, you are very welcome to participate in the course.