It looks like problems you get contain random parameters so students get slightly different problems (and the slightly different answers). At the moment we (instructors) are not completely aware about all features, bugs and quirks of this system. So, the question "which answer is correct?" or "is it a correct answer" makes no sense without writing down the problem itself.
Remember: system is experimental, beta-quality and we do not assign marks for Home Assignments.
