From bf53e86104aa8dbeb4543b54ef5894b77f020058 Mon Sep 17 00:00:00 2001
From: Mike Bourgeous <mike-bourgeous@users.noreply.github.com>
Date: Fri, 11 Oct 2024 22:58:17 -0600
Subject: [PATCH] Correct possible typo: source to bin

It looks like this sentence is trying to say that the energy of a source within a *bin* determines source domination of that bin, rather than the energy of the *source* (as is, "if most of the energy in that source is from [that source]" seems not to make sense, because *all* of the energy from a source would come from that source).
---
 book/approaches/deep/architectures.md | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/book/approaches/deep/architectures.md b/book/approaches/deep/architectures.md
index fd356ce..19c96c6 100644
--- a/book/approaches/deep/architectures.md
+++ b/book/approaches/deep/architectures.md
@@ -173,7 +173,7 @@ Deep Clustering maps each TF bin to a high-dimensional
 embedding space such that TF bins dominated by the same source
 are close and those dominated by different sources are far apart.
 We say that a TF bin is _dominated_ by some Source $S_i$ if most
-of the energy in that source is from $S_i$.
+of the energy in that bin is from $S_i$.
 
 Deep Clustering has the same basic network architecture as a
 Mask Inference network: spectrogram input, to batch norm, to a