I've only just realized - the negative variant doesn't break the pattern at all! In fact, -1,-1 actually looks like that and therefore is what you get from the positive integer limit. Insane!!
It is the opposite for true negative inputs, as in the negative integer limit. However, that one ends in an 8, which I believe is the reason it breaks the pattern by displaying 0 instead of 1? What's interesting is that it actually outputs 0,0 instead of diverging into the positives.
So, basically, really big positive numbers actually start to display as different numbers, like the 64-bit signed integer limit (positive) displaying as -1, -1 despite being a whole different number. The numbers diverge into another sequence and tend to actually get smaller and smaller, which can be proven by the 64-bit limit being displayed as -1. They get so stupendously small that it overrides TPT's code and diverge into the negatives.
I've only just realized - the negative variant doesn't break the pattern at all! In fact, -1,-1 actually looks like that and therefore is what you get from the positive integer limit. Insane!!