Instructions for External PhD Students

Here are my instructions on external PhD students:

Getting started as an external PhD student

I select a small number of external PhD student and I do want to highlight that it is a privilege to be selected as one. I only select students whom I consider outstanding and who have a day to day advisor whom I trust 100%. So, well done so far!

However, it directly comes with a few major expectations:

I want to be involved in your work and I expect you to make sure that I am involved. I do not serve as a PhD advisor if I am not part of the ongoing research. We need regular meetings - even if they are not as frequent as the meetings of the internal students.
We expect the same performance as for an internal student! We have high expectations on research quality.
Please try to collaborate with internal IAS members. So check out their homepages and get them involved in your research. Research is a team sport and you really have an excellent group of people to collaborate with here at IAS.

As a first step, I suggest you write a short research plan for yourself. As you may know, no research plan survives the first contact with the subject of study - but having no plan at the start guarantees that you will end up in a dead end. On the other hand, I can give you much better long term advice when we have your personal initial plan as a starting point for our discussions.

Make sure to read the following texts:

https://www.ias.informatik.tu-darmstadt.de/Towards-Finishing-Up/HowToGetAPhD?

https://www.ias.informatik.tu-darmstadt.de/Getting-Started/StartUpPackage

BTW: Also, give at least one talk per year in the Oberseminar...

Administrative Hurdles

Step 1: Admission is at the department level. Right at the beginning of your Ph.D. we should get you admitted - please fill and sign

https://www.informatik.tu-darmstadt.de/media/informatik/fb20_dekanat/dokumente_promotion/Annahme_als_Doktorandin_Doktorand.pdf

and send it to office@ias.tu-darmstadt.de where our fine secretaries will take care of the rest.

Step 2: Registration at university level. Once that is completed, we need to officially fill

a TUCAN application

Quals:

Once you will be two years with us, I expect, in theory, you should be half way with your PhD and could graduate in another one to two years. Thus, I think sometime in the second half of this year, you should come up for your “quals”. That means, I want you to write a preliminary thesis, copy the current journal papers into it and outline a plan for the remainder of your PhD. See:

https://www.ias.informatik.tu-darmstadt.de/Towards-Finishing-Up/PhDThesisProposal

Here is my PhD proposal (yes, I am ashamed of it, it should be a lower bound at best and the thesis became better) for reference:

https://www.ias.informatik.tu-darmstadt.de/uploads/Research/ThesisProposal/ThesisProposal.pdf

Also look at the theses of our recent PhD graduates! That should really help!

For the quals paper, I expect you to:

prepare all thesis parts that you can already prepare (e.g., abstract, Ch. 1: Introduction,

tables of symbols and abbreviations, your CV, your planned+current publication list, …)

Use \include{JournalPaper1.tex} for including your journal paper drafts. Don’t smoothen

things yet. If you don’t have journal papers

Instead of Ch. 5: Conclusion add Ch. 5: Plans until Graduation (which can be 1 page)

Past theses give a good guideline. Please don’t create any redundant or unnecessary text or elements except for “Ch. 5: Plans until Graduation.”

I strongly recommend to keep a file “Future_Work.tex” around where you “dump” all unfinished ideas. You could even put it into the Appendix of the Quals paper and later add it to the “Ch. 5: Conclusion” in the thesis.

The “quals” may be Qualen (German for huge pains) as we simulate a defense with the postdocs posing as committee members. But it *really* helps - you will know where your weaknesses and accomplishments are. Typically, your output will improve substantially.

So fair warning when you are half way (e.g. 18-24 months after starting and 18-24 before finishing), I expect you to have these three things ready:

the quals paper
a time plan for the second half
ideas for your external committee member
a quals talk = a test defense

We aim at making the quals hard such that the defense becomes easy!

Graduation:

I don’t want you to write a book which no one reads. I don’t want you to maximize the number of papers. I also don’t want you to get your first outside feedback. Nevertheless, I *DO* want you to have a high quality thesis. That means each meat thesis chapter should have the quality to be publishable in a reasonably quality journal. I will *require* that all meat thesis chapters to have been submitted to journals before you *defend*. That’s my absolute gold standard for theses.

I *recommend* that you have at least one accepted (more is better: I can argue better for a good grade, we can tell critical thesis committee “fuck off, three experts in the field have signed off on this chapter!”, you have the maximal readership & reward for your thesis chapters, and you have smoothened out all problems and typos). Most importantly, it makes sure that I have read your thesis chapters *LONG* before they become a thesis and that we are collaborators instead of adversaries during your defense.

BTW: Often journal papers will only be printed during your postdoctoral life - which can help you bridge the post-PhD publication gap by a paper…

I recommend attending a few defenses before yours. You will notice that I will ask the meanest question. I do that on purpose: After the defense, I can subsequently argue for better grades without anybody getting suspicious… :):)

Here is a summary of the steps involved in graduation:

https://www.ias.informatik.tu-darmstadt.de/Towards-Finishing-Up/Graduation