Reduced Graph States: Characterizing Basis States

In summary: A##, since these are the positions where there are 1s in the string. This results in a basis state that is orthogonal to all other basis states with different strings ##\mathbf{\Gamma} B'##, and they span a subspace of ##\mathbf{H}^{A}##.The motivation for choosing this basis is to have a set of states that are easy to manipulate and analyze mathematically, and that reflect the structure of the graph in a meaningful way. By defining the basis states in this way, we can easily see how the reduced state ##\rho_{G}^{A}## can be written as a sum of projections onto
  • #1
Johny Boy
10
0
TL;DR Summary
I have a paper query regarding a fundamental result on graph states, more specifically a result that characterizes a reduced graph state in a particular basis.
Reduced graph states are characterized as follows from page 46 of this paper:

Proposition:
Let ##A \subseteq V## be a subset of vertices for a graph ##G = (V,E)## and ##B = V\setminus A## the corresponding complement in ##V##. The reduced state ##\rho_{G}^{A}:= tr_{B}(|G\rangle\langle G|)## is given by $$\rho^{A}_{G} = \frac{1}{2^{|A|}}\sum_{\sigma \in \mathcal{S_{A}}}\sigma,~~~~~~~~~~~~~~~~~~(1)$$where ##\mathcal{S}_{A}:=\{ \sigma \in \mathcal{S}| \text{supp}(\sigma) \subseteq A \}## denotes the subgroup of stabilizer elements ##\sigma \in \mathcal{S}## for ##|G\rangle## with support on the set of vertices within ##A##. ##\rho_{G}^{A}## is up to some factor a projection, i.e.

$$(\rho_{G}^{A})^2 = \frac{|\mathcal{S}_{A}|}{2^{|A|}}\rho_{G}^{A}~~~~~~~~~~~~~~~~~~~(2)$$It projects onto the subspace in ##\mathbf{H}^{A}## spanned by the vectors $$|\mathbf{\Gamma}'B'\rangle_{G[A]} = \sigma_{z}^{\mathbf{\Gamma}'B'}|G[A]\rangle~~~~~~~(B' \subseteq B)~~~~~~~~~(3)$$where ##G[A] = G\setminus B## is the subgraph of ##G## induced by ##A## and ##\mathbf{\Gamma}':=\mathbf{\Gamma}^{AB}## denotes the ##|A| \times |B|-##off diagonal sub-matrix of the adjacency matrix ##\mathbf{\Gamma}## for ##G## that represents the edges between ##A## and ##B##:
$$
\begin{align}
\begin{pmatrix}
\mathbf{\Gamma}_{A} & \mathbf{\Gamma}_{AB} \\
\mathbf{\Gamma}^{T}_{AB} & \mathbf{\Gamma}_{B}
\end{pmatrix} = \mathbf{\Gamma}.
\end{align}
$$
In this basis, ##\rho_{G}^{A}## can be written as $$\rho_{G}^{A} = \frac{1}{2^{|B|}}\sum_{B' \subseteq B}| \mathbf{\Gamma}' B' \rangle_{G[A]} \langle \mathbf{\Gamma}'B'|.~~~~~~~~~~~~~(4)$$
Question:
The results of equation (1) and (2) are understood. I'm trying to understand the motivation for defining the basis states ##| \mathbf{\Gamma}' B' \rangle_{G[A]}## as shown in equation (3). As I understand, ##\mathbf{\Gamma} B' ## in the exponent of equation (3), is some string in ##\{0,1\}^{|B'|}##. In this way they would show that there are sufficient permutations of ##\mathbf{\Gamma}' B'##, where ##B' \subseteq B##, to produce orthogonal states ##| \mathbf{\Gamma}' B' \rangle_{G[A]}## which spans a subspace of ##\mathbf{H}^{A} \subseteq (\mathbb{C})^V##. Explicitly how is the term '##| \mathbf{\Gamma}' B' \rangle_{G[A]}##' defined? I don't really understand why the exponent is chosen as ##\mathbf{\Gamma} B'## into begin with to characterize the basis states?

Thanks for any assistance.
 
Physics news on Phys.org
  • #2

Thank you for your question. The basis states ##| \mathbf{\Gamma}' B' \rangle_{G[A]}## are defined as follows:

First, we take the subgraph ##G[A] = G\setminus B##, which is the induced subgraph of ##G## by the vertices in ##A##. This means that we only consider the edges that connect vertices within ##A##, and we disregard any edges that connect a vertex in ##A## to a vertex in ##B##.

Next, we consider the adjacency matrix ##\mathbf{\Gamma}## for the full graph ##G##. This is a square matrix with dimension equal to the number of vertices in ##G##. The submatrix ##\mathbf{\Gamma}'## is then defined as the ##|A| \times |B|## off-diagonal submatrix of ##\mathbf{\Gamma}##, which contains only the entries corresponding to the edges between vertices in ##A## and vertices in ##B##. This is why we use the notation ##\mathbf{\Gamma}'^{AB}## to denote this submatrix.

Now, for a given subset ##B' \subseteq B##, we can define the string ##\mathbf{\Gamma} B'## as the string of bits representing the presence or absence of edges between vertices in ##A## and vertices in ##B'##. For example, if we have four vertices in ##A## and three vertices in ##B##, the string ##\mathbf{\Gamma} B'## might look like ##011010##, where the first two bits correspond to edges between the first two vertices in ##A## and the first vertex in ##B##, the third and fourth bits correspond to edges between the third and fourth vertices in ##A## and the second vertex in ##B##, and so on.

Finally, the basis states ##| \mathbf{\Gamma}' B' \rangle_{G[A]}## are defined as the states obtained by applying the ##\sigma_{z}^{\mathbf{\Gamma}' B'}## operator to the state ##|G[A]\rangle##. This operator flips the sign of the state whenever there is a 1 in the corresponding position in the string ##\mathbf{\Gamma} B'##. For example, in the string ##011010##, the operator would flip the sign of the
 

Similar threads

Replies
11
Views
1K
Replies
10
Views
2K
Replies
1
Views
700
Replies
6
Views
1K
Replies
1
Views
2K
Back
Top