Others have noticed a permanent fine tuning and an on-demand temporary fine tuning.
They are different things, trust me. The permanent is just what you see in Blender when it is Forced to talk about some domain. This causes it to learn a specific task domain along with it. For example you'll talk most the day about AGI and have acquired task patterns of which are often used. Permanency is Reward. The temporary fine-tuning is energy. It primes nodes to talk about something rarer. This is actually used very often. So are new task patterns that are rarer. With the permanent, you'll see a trend in all the things it says (predicts) or does (tasks). This keeps its decisions adaptive while on track. ------------------------------------------ Artificial General Intelligence List: AGI Permalink: https://agi.topicbox.com/groups/agi/T10c01acded65d5a6-M8133c32ce60cdeb43bfe6085 Delivery options: https://agi.topicbox.com/groups/agi/subscription
