Setting the temperature to 0 mathematically tells the system to always choose the absolute highest-probability word (known as "greedy decoding"), but in no reality is this "deterministic". Output drift is still a thing.

But that's from, like, floating point errors, right? If you used higher precision that wouldn't happen, it's just because we're cheap in how we do rounding.

I can see how you'd nitpick this, but to me this is a deterministic algorithm that just happens to be running on nondeterministic hardware.