From a9abab5bfefd2cf297ae5e4ce725666dad29b0ba Mon Sep 17 00:00:00 2001
From: Matthijs Douze <matthijs@fb.com>
Date: Fri, 1 Sep 2023 18:04:49 +0200
Subject: [PATCH] Update README.md

---
 neurips23/filter/faiss/README.md | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/neurips23/filter/faiss/README.md b/neurips23/filter/faiss/README.md
index 13cbfaaf..c834af51 100644
--- a/neurips23/filter/faiss/README.md
+++ b/neurips23/filter/faiss/README.md
@@ -92,6 +92,9 @@ Therefore, the rule that we use is based on a threshold $\tau$ (called `metadata
 
 - otherwise use the IVFFlat based index
 
+Note that the optimal threshold also depends on the target accuracy (since the IVFFlat is not exact, when a higher accuracy is desired), see https://github.com/harsha-simhadri/big-ann-benchmarks/pull/105#issuecomment-1539842223 .
+
+
 ## Code layout 
 
 The code is in faiss.py, with performance critical parts implemented in C++ and wrapped with SWIG in `bow_id_selector.swig`.