1: \begin{definition}[Normalized neural persistence]

2:   For a layer $G_k$ following Definition~\ref{def:Layer}, using the

3:   upper bound of Theorem~\ref{thm:Neural persistence bounds}, the

4:   \emph{normalized neural persistence} $\normneuralpersistence(G_k)$ is

5:   defined as the neural persistence of $G_k$ divided by its upper bound,

6:   i.e.\

7:   %

8:   $\normneuralpersistence(G_k) := \neuralpersistence(G_k) \cdot \neuralpersistence(G_k^+)^{-1}$.

9: \end{definition}

10: