I had a hard time finding details about research scientist interviews, so here's my single round experience. There were two back-to-back interviews.

1. Discussion + coding with research scientist
   • Questions about research area, going from shallow to deep.
   • A few basic ML questions (e.g., what is the logistic regression cost function?).
   • Coding: Find the edit distance between a pair of strings. (I was told this was a basic question in signal processing, which is not my subarea.)
2. Coding with research/software engineer
   • Given a binary (obstacle or not) n x m matrix, count the number of paths from upper-left corner to bottom-right. What is the precise size of the search space?
   • Implement linear regression model from scratch in numpy.

I was very ill-prepared for this interview as it was the first tech round I had for a research-related position. The recruiter gave me almost no details besides saying that there would be some coding and ML. If you interview with them, I'd recommend brushing up on the basics of ML theory if you haven't looked at it in a while. No coasting by on your research experience.