From 5127bc88f077946206f4bf61b148b28c625c1926 Mon Sep 17 00:00:00 2001
From: Xianda Sun <5433119+sunxd3@users.noreply.github.com>
Date: Thu, 17 Oct 2024 18:41:55 +0800
Subject: [PATCH] Update design_notes/interface_for_gibbs.md

---
 design_notes/interface_for_gibbs.md | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/design_notes/interface_for_gibbs.md b/design_notes/interface_for_gibbs.md
index e776850..2cbdf1e 100644
--- a/design_notes/interface_for_gibbs.md
+++ b/design_notes/interface_for_gibbs.md
@@ -15,7 +15,7 @@ The goal is to document some of the considerations that went into the closed PR
 
 ### Recomputing Log Densities for Parameter Groups
 
-Let's consider splitting the model parameters into three groups (assuming the grouping stays fixed between iterations). Each parameter group will have a corresponding sampler state (along with the sampler used for that group).
+Let's consider splitting the model parameters into several groups (assuming the grouping stays fixed between iterations). Each parameter group will have a corresponding sampler state (along with the sampler used for that group).
 
 In the general case, the log densities stored in the states will be incorrect at the time of sampling each group. This is because the values of the other two parameter groups can change from when the current log density was computed, as they get updated within the Gibbs sweep.