How to verify widening conversion?

[Gribouillis]

If the widening is not present in the bytecode - is there another way I can verify it?

By widening, I suppose you mean the conversion from int to float that may occur when the expression 5.0 == 5 is evaluated. In CPython, this is done in the C function float_richcompare() of Objects/floatobject.c in the source tree, see the section that starts with else if (PyLong_Check(w)) {.... This part is rather tricky because Python integers can be arbitrary large.