A0 subtlety

[jackd]

Was a set of input planes representing a position at time (t - T + 1) oriented for the side to move at time t or time (t - T + 1)?

[Daniel Shawul]

Was a set of input planes representing a position at time (t - T + 1) oriented for the side to move at time t or time (t - T + 1)?

All the history input planes are oriented for the current side to move, i.e. at time t.

[chrisw]

Was a set of input planes representing a position at time (t - T + 1) oriented for the side to move at time t or time (t - T + 1)?

All the history input planes are oriented for the current side to move, i.e. at time t.

Would it matter if they weren’t?
(That’s a serious question btw)

[Daniel Shawul]

Was a set of input planes representing a position at time (t - T + 1) oriented for the side to move at time t or time (t - T + 1)?

All the history input planes are oriented for the current side to move, i.e. at time t.

Would it matter if they weren’t?
(That’s a serious question btw)

I suppose not but it would make it a little bit harder because the network has to learn to flip the board
every step in the history. I do not think history planes matter much anyway to begin with.

[chrisw]

Was a set of input planes representing a position at time (t - T + 1) oriented for the side to move at time t or time (t - T + 1)?

All the history input planes are oriented for the current side to move, i.e. at time t.

Would it matter if they weren’t?
(That’s a serious question btw)

I suppose not but it would make it a little bit harder because the network has to learn to flip the board
every step in the history. I do not think history planes matter much anyway to begin with.

I thought it would depend on how the convolution filters were tracking the data (in what order). Does it track 3x3 across each 8x8 map of chessboard cells, and then combine those outputs - so I suppose it must be best to combine a1 to a1 region and so on, rather than a1 to a8. Difficult to wrap head around.

[AlvaroBegue]

Was a set of input planes representing a position at time (t - T + 1) oriented for the side to move at time t or time (t - T + 1)?

All the history input planes are oriented for the current side to move, i.e. at time t.

Would it matter if they weren’t?
(That’s a serious question btw)

I suppose not but it would make it a little bit harder because the network has to learn to flip the board
every step in the history. I do not think history planes matter much anyway to begin with.

I thought it would depend on how the convolution filters were tracking the data (in what order). Does it track 3x3 across each 8x8 map of chessboard cells, and then combine those outputs - so I suppose it must be best to combine a1 to a1 region and so on, rather than a1 to a8. Difficult to wrap head around.

I agree with this. However, it's very likely that the history inputs are mostly irrelevant, as Daniel Shawul mentioned.

[chrisw]

Was a set of input planes representing a position at time (t - T + 1) oriented for the side to move at time t or time (t - T + 1)?

All the history input planes are oriented for the current side to move, i.e. at time t.

Would it matter if they weren’t?
(That’s a serious question btw)

I suppose not but it would make it a little bit harder because the network has to learn to flip the board
every step in the history. I do not think history planes matter much anyway to begin with.

I thought it would depend on how the convolution filters were tracking the data (in what order). Does it track 3x3 across each 8x8 map of chessboard cells, and then combine those outputs - so I suppose it must be best to combine a1 to a1 region and so on, rather than a1 to a8. Difficult to wrap head around.

I agree with this. However, it's very likely that the history inputs are mostly irrelevant, as Daniel Shawul mentioned.

well, for me, it is clear that they, or some form of equivalence, are vital. Attempts at explanations have met with uniform blank looks or raised eyebrows, so I’ll just stick with this view, and we’ll see. It won’t be the first time.

[jackd]

If the input is always oriented for the side to move, then should the policy output also be mirrored for black? Like if the best move was g8f6 the policy output would b1c3?

[Daniel Shawul]

If the input is always oriented for the side to move, then should the policy output also be mirrored for black? Like if the best move was g8f6 the policy output would b1c3?

I always evaluate for white to move, so the board has to be flipped horizontally when it is black to move and ofcourse all moves are flipped as well
So g8f6 becomes g1f3 (not b1c3). Btw this has nothing to do with your original question of history planes.

@chrisw Good point on the a1->a8 mapping. Indeed the innder dot product over the volume (channels) will be messed up if the history planes are not aligned properly.

[jackd]

@Daniel Shawul

How strong is your policy network on it's own?