As best I can recall, yes - that is / has been the prevailing wisdom on this topic. For the $EE mask (again, as best I can recall) it's believed that this table contains signed integers that are all negative.

Just for a sanity check though, can you verify the address of the table you're talking about pls? I've been swamped at work for the last couple weeks and simply haven't had time to delve into tuning stuff.