Similar here. I like being around other people to some degree, and it can be energizing in some ways, but it also uses energy.
It also uses attention, and maybe engages different kinds of thinking, so I can't focus as deeply on hard problems.
One of my programming styles, for example, is "meticulous zero-defect and resilient", and when someone is in social proximity, it gets noticeably much harder, to impossible. I can do a routine UI design and coding when someone is right next to me, but even then, don't expect me to flash on a brilliant insight into creative UI design or application opportunity -- only to do something that I don't have to think about much.
(This is also one of the many problems with Leetcode hazings/shibboleths -- if you're trying to think, rather than just instantiate the interview ritual you've memorized and practiced so much, that you can do it in your sleep.)
(A related poor awareness of interviewers is that I'm pretty sure many don't realize how much signal interviewers can pickup on intuitively, from language, voice, and body language. That note at the start of the syllable in the middle of innocuous-sounding sentence the Leetcode whip-cracker just said... can tell some interviewers quite a lot about what the person is really thinking, or can even be a concerning flag about their personality. Which can be very distracting when you're thinking about an algorithm that suddenly doesn't seem like the biggest problem at the moment. And if you have to also break your focused thinking to keep up a real-time expository narrative of your process for interview ritual purposes (which isn't going to match actual cognition, so extra load)... careful not to start also narrating how you simultaneously need to figure out how to gently show them something they seem to misunderstand, without risking an insecure backlash when they do the interview writeup, because they're coming across like that might be a risk.)