From Floor Pi Wiki
Jump to: navigation, search
Brian Chen
Room H302
Year 2019
Courses Mathematics, Electrical Engineering and Computer Science

Hi. I'm Brian. I currently live in H302. I no longer live in H302, or on floorpi for that matter. I am from Taiwan but know an alarming number of other people on floorpi from AoPS. I like math, computer science, singing, puzzles, dragons, and lots of other stuff.

Ooh, an infobox! (My courses are only tentative; I have an open mind with regards to changing or dropping majors. But it looks cool!) I think it is pretty safe to say that my courses are pretty locked down, since I've graduated with them and all.

I was a hall chair Fall 2017–2018 (with Jakob) and also sort of HistoryComm. Wao people update their wikis.

You can look at my quasiprofessional site here:


Fall 2015

  • 5.111, Principles of Chemical Science (van Voorhis, Shoulders): Well, it's a GIR. I liked the lectures and lecturers; they kept the whole thing at least somewhat engaging. Other parts, like the online problems, were decidedly meh.
  • 6.046J/18.410J, Design and Analysis of Algorithms (Moshkovitz, Goldwasser, Shavit): Coming from an extensive competitive programming background, I probably don't have any thoughts or opinions on this class that would be helpful to others.
  • 8.022, Physics II (Electricity and Magnetism) (Ashoori): The professor is a cool guy and there are cool experiments, but the hardest parts of lecture consist of a sea of equations and it was hard for me to to see the forest for the trees often. On the other hand, my TA (Zwierlein) was super awesome.
  • 17.309J/ESD.082J/STS.082J, Science, Technology, and Public Policy (Oye; McDonnell): A CI course I have to keep spelling out. It's pretty intense; there are a lot of readings, two 10--15-page essays, and a midterm and final from an essay pool. The reward is the professor intersperses awesome stories to tell throughout his lectures. The guy really gets around and never runs out of relevant tales of how he met so-and-so powerful political figure on an airplane, or gave a presentation to a room full of company CEOs, or whatever. I'm quite glad I took it, but I'd say you shouldn't take it unless you're sure you're interested in the material.
  • 18.A34, Math Problem Solving (aka "Putnam Seminar") (Shor): Well, we do what it sounds like, solve vaguely Putnamesque problems but with lots of other weird things on the side.

Spring 2016

  • 6.01 (hartz?): To be generous, I learned about op-amps. The other stuff was routine. Don't go in without knowing Python; you'll struggle and it won't be your fault.
  • 6.841J/18.405J (Moshkovitz): Complexity theory is amazing. Unfortunately Moshkovitz is leaving and I don't know how much useful information I can still give about this class.
  • 18.702 (Negut): I got it the semester Artin decided not to teach it, and it was taught by Negut, who was too fast for the first bit of the class, but slowed down soon enough. He also has an unfortunate stutter, but I would still say a decent lecturer.
  • 21M.011 (Pollock; Neff): A nice CI class in that the papers are mostly about careful listening and giving an informed opinion, with minimal research. The listening quizzes can be hard, but if you like music this is pretty good. Pollock is a dedicated lecture and Neff is a fantastic section leader.
  • 21M.301 (Ruehr): 10/10 would take again. I was in Ruehr's section. The homework was very light, but I learned a lot and the final project, where you get to compose a short string quartet piece and hear it performed for you, is really cool. There's a piano lab and sight-singing lab, the latter of which was quite strict with attendance, but that's all.

Fall 2016

  • 6.004 (Terman?): You can flexibly frontload lots of work. The design project is fun. I didn't go to a single lecture, but that's probably just me.
  • 6.867 (Kaelbling, Sra, ?): I don't get machine learning, it seems like you just throw lots of things onto a metaphorical table and jiggle it until you get something good. The tests are weirdly easy algebraic term-pushing. The projects are sort of interesting and really let you try and implement the concepts/algorithms (or call outside libraries to do them). Do make sure you know numpy or MATLAB; optimally, have a partner who knows the same system. I think numpy is a safer bet, since it's easy to use TensorFlow when that becomes useful in neural nets, as well as scikit-learn and others.
  • 9.01 (Bear, Lambo): Wow, all quizzes and tests, no p-sets. The lecturers are standard but do have a sense of humor. There is some rote memorization (lots of Latin names for parts of the brain), but not as much as I was concerned, it wasn't weighted that heavily, and the lectures emphasize what things are important repeatedly (you learn names so you can use them down the road). I'm fascinated by learning about my own brain.
  • 18.217 (Stanley): We just went over Enumerative Combinatorics. It's kind of slow, but I probably still didn't listen as much as I should have. Well, I like combinatorics... The later lectures developed a lot of theory about polynomials and power series that just never clicked for me, personally. I don't know if this class is going to be taught much more, though (is Stanley retiring?) so it might not end up mattering.
  • 6.854 (Karger and Madry): Decent lecturers but the p-sets are rapid and intense. External circumstances for me made the final project a lot more of a slog than it could have been; I did a relatively unadventurous reading project.

Spring 2017

  • 6.033 (LaCurts; Szolovits, Carleton) Katrina is a fine lecturer. Peter is kinda slow but full of stories and nice. Amy is OK and held a write-together party with snacks. Being a CI-M and all, this class is what it is.
  • 7.013 (Sive, Amon; TA Navarro): Well, it's also a GIR. Lectures are fine. I had a really good TA.
  • 18.901 (Walpulski): Lectures are pretty standard. This class is nontrivial, but not that nontrivial. Having read the little topology in Rudin let me coast through maybe half the class lectures, although some (very few) of the exercises were very interesting.
  • 21M.250 (Neff): Neff is a boss.

Summer 2017

I did SPUR. My paper is honestly kind of silly.

Fall 2017

  • 6.035 (Rinard, Carbin): I took this class because I was roped into an entirely Floorpi-based team. Compilers are a lot of work idk. Rinard is a bit much in his lectures for me, think sort of like an educational TV program for elementary school students where they repeat things in excited voices plus snarky sage life advice except it's compilers, but some others really enjoy him. Carbin is standard and interactive to a reasonable amount. Lecture schedule is weird and randomly pauses some weeks.
  • 6.805/STS.082 (Weitzner, Edelman, Fischer): I think internet policy is really interesting and the profs are experienced and good at lecturing and know their stuff. The CI part is not bolted on, which is good. (Note that this is a CI-M I have no chance of using as such; I took it as a HASS.)
  • 18.785 (Sutherland): I think this is easily the hardest class I've taken at MIT, but I didn't have the corequisite 18.705 (people told me not to take it because it was slow) or a lot of algebra background in general. Somehow the psets got easier later.
  • 21M.385/6.809 (Egozy): This class is way cool, but hard to get into; I am glad I made it in on my first try. The professor is one of the cofounders of Harmonix, best known for making Guitar Hero (but the brand got sold to Activision so then they made Rock Band).

I switched from 18C to the all-too-common 18 + 6-3. I also did a UROP that's sort of related to my SPUR project but with m a c h i n e   l e a r n i n g.

Spring 2018

  • 6.831 (grad version of 6.813) (Karger, Zhang, Verou; Friedman) My 6.035 floorpi team continued as "necrodancers". This really commits to the whole flipped lecture thing, you actually have to do readings before class and then go to class to reinforce and practice what you learned (and also get nanoquizzed). But it also does it pretty well, without being onerous.
  • 9.00 (Gabrieli; Ayyash) This overlapped more with 9.01 and/or AP Psychology than I expected. However, lectures are pretty engaging. There are a lot of demonstrations. The professor takes steps to make the material interesting even if you're not super interested in psychology. Since I am, it was better.
  • 18.200 (Moitra, Goemans) Lalala I need CI-Ms! My advisor warned me I would likely have seen most of the material before, but I didn't process that he would be one of the lecturers until after I showed up at class. But I did learn about writing math, more so than I expected. In some sense it was just making explicit things I vaguely knew in the back of my head, but that is useful.
  • 21L.430/CMS.920 (Donaldson) I, uh, have feelings.
  • Spark director (this was probably more work than any of my classes)

Fall 2018

  • 6.004 LA: Bluespec is a thing.
  • 6.UAT (Eng; Friedman + oh shoot I don't remember his last name): I have no idea what the title even is, but this class is specifically and generally about public speaking, and is nothing short of lifechanging for some amount of time. You are forced to give all sorts of presentations from all across the gamut and are forced to the edges of your comfort zone. It's pain and suffering, but you come out a much better presenter and will never look at presentations the same way again. Would recommend, but if it was taught as it was this year, the timing requires a bit of thought: retroactively I would pick junior fall so that one has enough generalized background with technical projects to have content to present about, but still lots of chances to apply the skills in other classes.
  • 18.434 (Remscrim): CI-M. Nothing special. Weirdly we ended up doing spectral graph theory. CS applications come later I guess. I gave two blackboard talks on sections of the textbook and one presentation with slides that was kind of terrible because MIT wifi decided to go down or something.
  • 21L.489 (Interactive Narrative) (Montfort): Nick finds the craziest music videos to analyze. Fun except for how I am terrible at creative writing (??). That was also pain and suffering but I am very happy with what I ended up producing. Legitimately the only class this semester in which I wrote any code (Inform 7).
  • 21M.401 (Concert Choir) (Cutter): Fun. I don't know why we have three concerts this semester. I got to sing for Jacob Collier and watch him give a concert in pajamas and no shoes, but that was probably just blind luck.

Spring 2019

I'm too jaded to give proper reviews now.

  • 6.004 LA: Bluespec is still a thing.
  • 6.857 (Rivest, Kalai): Writing the lyrics to "Never Gonna Give You Up" on the blockchain was fun.
  • 6.858 (Kaashoek, Morris): Hacking stuff was kinda fun but pretty handholdy since I already do CTFs.
  • 21M.302 (Shadle): I got to write a theme and variations.
  • 21M.401 (Cutter): Continued from last semester. This was a normal semester where we just had one concert. I dunno, it was still fun.

Fall 2019

  • 6.004 TA: help
  • 6.826: coq is fun!

Spring 2020

  • 6.004 TA: after the disaster of fall 2019, there was nowhere to go but up
  • 6.822: coq is still fun! adam's tactics are pro

Not Classes

  • I like puzzles. Remotely, I hunted with Random twice and wrote a hunt with them once before coming to MIT. Once here, I participated three puzzle hunts (Simmons REX hunt, aquarium hunt, Mark Halpin's 2015 Labor Day Extravaganza) before classes started. Throughout the year I hunt with floorpi on Australian hunts and the occasional smaller local hunt, and ✈✈✈Galactic Trendsetters✈✈✈ on Mystery Hunt. I've written puzzles with the Trendsetters for the 2017, 2018, and 2019 Galactic Puzzle Hunts.
  • I like weird songs. I know the Pathery theme song by heart because it exists on this wiki, even though it's not a thing any more. I performed "We Didn't Start the Fire" and the Animaniacs' "The Names of All 50 States and their Capitals" at Open Mic Night 2017. Jonathan Coulton's "Shop Vac" has an amazing kinetic typography video.
  • I am super involved in the Student Information Processing Board (SIPB, W20-557) and the Educational Studies Program (ESP, W20-467) so spend a lot of time in W20. Talk to me if you want to get involved. We have cool places to hang out, power, food, and free t-shirts. Honestly like talk to me if you just want free t-shirts.
  • I did lots of mathematics and programming competitions in high school, and kept doing some of them here — Putnam, ACM-ICPC. But you might have already guessed that? This was mostly my first two years; I got hosed with other things afterwards.
  • I wrote a Castlefall client for floorpi. I also flexboxed/gridded the main page while procrastinating during spring break.

Here's a Triple Back (from my blog, prior example and rules from MellowMelon).