2 ms·
You can assume the "computer" solution is overfitted, too. There's no reason to, because general methods are well-known, but it's even easier to just hardcode a
by thaumasiotes 7y ago
You can assume the "computer" solution is overfitted, too. There's no reason to, because general methods are well-known, but it's even easier to just hardcode a cube and the list of moves that solves it than it is to implement one of those general methods.
Why assume that "I solved a Rubik's cube using a neural network" guarantees that I cheated?