1
$\begingroup$

I have two questions:

1) I have been reading an article on AlphaGo and one sentence confused me a little bit, because I'm not sure what it exactly means. The article says:

AlphaGo Zero only uses the black and white stones from the Go board as its input, whereas previous versions of AlphaGo included a small number of hand-engineered features.

source: https://deepmind.com/blog/alphago-zero-learning-scratch/

So what exactly is the input here? What do they mean by "just white and black stones as input"? What kind of information is the neural network using? The position of the stones?

2) What type of neural network is AlphaGo Zero using?

$\endgroup$
  • $\begingroup$ Welcome to SE:AI! (Although there are two questions here, they are related.) $\endgroup$ – DukeZhou Apr 22 at 20:57
1
$\begingroup$

The input to the neural network is a $19 × 19 × 17$ image stack comprising $17$ binary feature planes. $8$ feature planes $X_t$ consist of binary values indicating the presence of the current player’s stones ($X^i_t = 1$ if intersection $i$ contains a stone of the player’s colour at time-step $t$; $0$ if the intersection is empty, contains an opponent stone, or if $t < 0$). A further $8$ feature planes, $Y_t$ , represent the corresponding features for the opponent’s stones. The final feature plane, $C$, represents the colour to play, and has a constant value of either $1$ if black is to play or $0$ if white is to play. These planes are concatenated together to give input features $s_t = [ X_t, Y_t, X_{t−1}, Y_{t−1}, ..., X_{t−7}, Y_{t−7}, C]$.

This and all the other architecture details can be found in the "Neural Network Architecture" section in the paper.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.