Meta Research Scientist interview questions

I interviewed with the neuromotor interfaces team. The overall experience was pleasant and speedy. I first spoke to a recruiter and then did a full loop interview across 2 days (1 technical, 1 design, 1 behavioral, 1 presentation, and 2 more 1:1 interviews). For the presentation, I spoke about my work in academia (no different than your standard conference/job talk). The technical was very difficult. The task was to take a physiological signal, and implement K-means clustering from scratch to try to separate the parts of the signal that derived from different hand movements. You were not explicitly told to use K-means, just some "clustering" method (but K means is the way to go).

I had five interviews overall. Two for coding, two for research (ML in my case) and one for HR interview. The coding interviews were not as hard as the ones SWE are having, which made a lot of sense to me. All of the questions were taken from different preparation sites (I saw most of the questions for the first time during the interview but I was able to locate them to see if I got those right). The ML interviews both were about designing an ML pipeline from understanding what is the basic need of this pipeline up to deploying it.