Dear TAs,
We have a question regarding the state representation of the TicTacToe game. In particular, in Part1 we are supposed to use an array 3x3x1 with values from {1,0,-1} while in Part2 it says to use a 3x3x2 representation with values from {0,1}. We were wondering, why exactly do we need to change the way we represent the state? Also, do we need to change the whole TicTacToeEnv class or can we simply reshape the input that we give to the neural network?
Thank you for your help and I wish you an awesome day.
Best,
Luca & Robert
