A panel-style interview was conducted, where the applicant must audit some sample source code in front of 7 or so people, all of whom are giving feedback. It honestly was a mostly ineffective interview, which was unnecessarily unnerving. The vast majority of security researchers in this industry prefer to work and think in quiet, calm settings, not in a group setting where over half a dozen other researchers are shoulder-surfing your work. Dumb.

2 rounds first one be able to explain vulnerabilities and mitigations. second one find vulnerabilities(different from generic buffer overflow) and also reverse engineer a few small binaries. brush up on ctf style questions