Oh. My algorithm does something similar to SSE. But I think it's a bit different. The priming I use considers how far back each character is in the last ~500 characters, and also use what you called MATCH but that's for longer than order 0 (and I do that for the first few orders too).
I have at least 1 more question for you though. How does you algorithm (if it does?) handle time delay matching? Here's an example of the problem: You've seen in the past "the large large large um um um cat chewed" and you see in the future "the cat ?_?" and can in theory use the old memory's prediction - but you need to match that old memory - how does your algorithm handle this or does it not use these? ------------------------------------------ Artificial General Intelligence List: AGI Permalink: https://agi.topicbox.com/groups/agi/Tf0bedfcd44454678-M6f20a68e15f8732a435c8007 Delivery options: https://agi.topicbox.com/groups/agi/subscription
