After I took a three-month internship at MPI-SWS in 2016, I was asked a lot about how that felt like. Generally speaking, it was great! – and this is also how I replied to most people. However, I really want to share more, and hopefully, attracting you to apply for a research intern or long term position there!

MPI-SWS and Saarbrücken

MPI-SWS is based in two small towns at Germany: Saarbrücken and Kaiserslautern, they are near the boundary with France. These two sites are basically equal in terms of faculty size, and most research group have people in both places. Because they are actually very close to each other (one hour’s bus), we visit another institute one day each week in turn. Many senior faculties have offices in both places actually.

MPI-SWS building at Saarbrücken

I was in Saarbrücken, so I am gonna to talk about it. Saarbrücken is a lovely town, despite small. Like many other Germany cities, it can be a bit awkward to go around if you don’t speak Germany, but most people are quite nice with foreigners, and almost all clerks in stores and staff in bank etc. can speak English.

One thing quite “Germany” that can affect your life is that most stores will close on Sunday, and banks don’t open on weekends (even for daily operation, banks tend to close earlier than the time you finished your day’s work). So you need to secure your next week’s groceries in case Sunday came and you found nothing in your refrigerator (don’t laugh, this was me in my first weekend and all I ate that day was fruit – so veggie).

But the institute, despite being under the MPI system and located in Germany, is very internationally focused, and we usually think it was modeled after U.S. academic system compared to other European schools. Just take a look at our faculty members, our PhD program, and our Dell computers. It now even has a joint PhD program with UMD as well, and you should apply for it!

As far as the living condition, the research intern position has a decent stipend. And there was institute-level housing arrangement, so you don’t have to worry about it either. I was living in the MPI guest house, just 3 min’s walk from the institute (conveniently for work (late) but unfortunately 20 min far from the nearest supermarket by bus, since the institute was located in Saarland University’s campus outside the town).


I was working at the PLV group in Saarbrücken, and my advisor is the amazing Derek Dreyer. For day-to-day work, I will collaborate and get hands-on help from PhD students (Ralf Jung, Hoang-Hai Dang, and Jan-Oliver Kaiser) and a PostDoc (Jacques-Henri Jourdan), sincere thanks to you all. The larger group, PLV @ Saarbrücken will have a standup meeting twice each week, reporting what you was working on in recent days, which is usually very short (in total < 30 mins and I can finish my part in three lines. But it can be very confusing at first because I were new plus not quite familiar with other topics in PLV). However, the smaller team working on the same project would discuss and meet a lot. It was much more casual and informal, because I can just show up at their office and try to get advice when they are not so busy. This is a very fast learning experience. I didn’t meet with the my advisor a lot excluding the standup meetings, but for once or twice a week, he will let me “drop by my office at 10am and let’s try to flesh out the details of this proof” etc.

Window over my desk

At my first day, I left my apartment at 9 am in a hurry nervously hoping that I were not too late, but just found that I were the first to arrive at my office room. Ok, maybe as a Chinese I am not supposed to be surprised (LOL), but during the entire internship, you can feel that there is really no hard requirement over your schedule. As long as your research is progressing, and you will arrive at your meeting appointment in time, you are good. So I can leave earlier to buy groceries, or stay until 11 pm to finish my work at hand. Actually, in the later stage of my internship, I usually left early, like 4 pm, to work out and cook dinner, then came back to office at 8 or 9 pm and worked until 11 pm (because of my irregular schedule, my French colleague thought that I didn’t sleep!). I sometimes worked at weekends as well, if I was not going out, but for most people in MPI-SWS, working in office at weekends can be really, really rare and the entire floor would have <= 3 people at weekends. I mean, I am not promoting this or proud of my work style. We should have a life, just that it was too luxurious a thing to have for me at that time.

Inside MPI-SWS

Finally, I didn’t really get to publish anything that was done in my internship, partly because we kinda agreed that it was not worthy (my advisor was quite rigid about the venue of publication – if you know PL research, you will understand that by checking out his publications). But my work were still a useful contribution towards the overall thing, at least I got acknowledged in a POPL paper + mentioned in a POPL conference talk – well that is not a big deal but “shrugging”. BTW, my bachelor’s thesis was partly based on the work of my MPI-SWS colleagues, though not directly related to my research done there.



We got a lot of activities inter- or intra-group, like eating lunch together at the Saarland University’s canteen, climbing each week, dinner plus bar hunt each weekend, visit to Kaiserslautern (including a coffee break + ping pong time + PL reading group + various meetings), institute retreat (bi-annually, I was lucky to be there). I knew A LOT OF people during that period, and learned a lot about them. Noting that many of them were super interesting, super smart, and some are already established researchers, so it was insanely rewarding.

Because of the two-site problem, we actually set up a digital “mirror”, a big display connecting two open spaces at both Saarbrücken and Kaiserslautern, in real time, audio included. At first, I have no idea what that buzzing thing is and thought that this is just a default channel routed to a monitor in the institute building accidentally – However, when I visited Kaiserslautern one day, I saw the same thing and asked someone until I knew that this is a part of the remote conferencing system. Quite cool!

I think MPI-SWS’s intern culture was quite unique, and there will be interns around the year (though more during the summer). The interns’ backgrounds range from junior undergrads to PhD students, coming from Europe/America/Asia, and all interns I talked to gave me an impression that they thought their project was reasonably challenging, interesting, and relevant to the big picture of the research group. During the time off, we interns usually went for dinner and drinks, a lot of burgers, cocktails, shots … what a lovely time.

Beer Garden


Regarding perks, not so much, but I was satisfied.

One thing that people complained a lot (half-jokingly) was that MPI-SWS didn’t provide free coffee (“MPI can readily pay for your conference trip to somewhere exotic on earth, like Iceland, but not free coffee”). Luckily, the nearby MPI-Informatik did, and sometimes after lunch, I would take my MPI-SWS mug, join the MPI-SWS mug squad, and go hijack MPI-Informatik’s coffee machine.

Another cool thing was that I joined the institute’s English reading group, which was designed for non-native-English speakers to train their communication and presentation skills. The instructor Rose was very nice. I got a lot of things corrected, esp. my wrong pronunciation of “usual” (I was saying it like “euro” at that time). A lot of thanks.

We also have a mail-list in which people would drop a line if they took some food to share. That happened too frequently and I didn’t even bother to take one cake or something unless I came by the kitchen for some other reason. And when you ate the cake, you can stare at our beloved institute-named-after:


Sum up

I think my adventure at MPI-SWS is really a determining factor in my choice of going after a PhD degree in future. Not just that I got a good letter, but that I see the value of research, and I experienced the life as a researcher. No matter where my actual PhD at UW will lead me to, I got to say:

Thanks, Derek.