diff --git a/src/learn/learn.h b/src/learn/learn.h
index 2ee2f8d6..6056e8c6 100644
--- a/src/learn/learn.h
+++ b/src/learn/learn.h
@@ -5,18 +5,6 @@
 
 #include <vector>
 
-// ----------------------
-// learning from zero vector
-// ----------------------
-
-// Start learning the evaluation function parameters from the zero vector.
-// Initialize to zero, generate a game, learn from zero vector,
-// Game generation → If you repeat learning, you will get parameters that do not depend on the professional game. (maybe)
-// (very time consuming)
-
-//#define RESET_TO_ZERO_VECTOR
-
-
 // ----------------------
 // Floating point for learning
 // ----------------------