These are definitely some kind of fixed test values and cheap hamming-distance maximization. 4 00000100 56 00111000 66 01000010 115 01110011 155 10011011 156 10011100 230 11100110 237 11101101 1) 6 out of 7 columns have "balanced bits": four "1"s 2) XOR of all of these is 1. Suppose instead these were generated at random from [1,255]. Out of a million trials, the chance of fulfilling conditions 1,2 simultaneously is less than 1/25000. Simulation: colab.research.google.com/dr…