@acomodi@antmicro.com Please put the output examples inside a single cell table.

Proving this for all seeds is likely infeasible. From looking at vtr_random.cpp, there can be about 2 billion possible seeds, so proving it exhaustively is expensive. Using formal methods would also be very difficult.

This may seem easier, but is also equally difficult, because we know nothing about the population of designs, so it is possible that the tool could fail on every design not yet tested.