Update utils_ncb.py

utils_ncb.py

@@ -307,7 +307,7 @@ def get_parser(device):
         "0 indicates we just take that slot with the maximum slot attention value,"
         "1 indicates we take the maximum count of high attn weights (based on thresh_attn_ob_slots), "
         "otherwise those slots that contain a number of values above thresh_attn_obj_slots are chosen"
-        "(see retrievalbinder for usage)",
+        "(see neural_concept_binder.py for usage)",
     )
 
     parser.add_argument("--device", default=device)