From the AlphaZero paper, the caption of Table S1 (p. 13)
Table S1: Input features used by AlphaZero in Go, Chess and Shogi respectively. The first set of features are repeated for each position in a $T = 8$-step history. Counts are represented by a single real-valued input; other input features are represented by a one-hot encoding using the specified number of binary input planes. The current player is denoted by P1 and the opponent by P2.
I wanted to get some clarification on which features are represented by single real-valued inputs, and which "other input features" are represented by one-hot encodings.
I'm also not sure what the line
other input features are represented by a one-hot encoding using the specified number of binary input planes
For Chess, specifically, we have the following features): P1 Piece, P2 Piece, Repetitions, Colour, Total move count, P1 Castling, P2 Castling, No-progress count.
I think I understand P1 Piece and P2 Piece, these are just 6 binary planes, one for each piece type, that denotes whether a piece exists for each of the 64 squares on the board. Correct me if I'm wrong.
Repetitions, Colour, total move count, P1 Castling, P2 Castling, and No-progress count - are these all represented by "single real-valued input"? For example, if it's White's turn to move, and white is represented as 1, does this mean the plane for color is an 8x8 matrix of 1s? Another example: if the move count is 114, then the input plane for this feature is an 8x8 matrix of 114s...is this correct?
Follow up question: why are there 2 planes for repetitions?