LTM Knowledge Storage Decision Loop evaluate no / transient task state durable value retrieve beliefs stored for later infer new store derived New Information Arrives Worth remembering? Discard / History Only PIN: Working Memory REMEMBER: LTM Store QUERY: LTM Retrieval NAL/PLN Reasoning Derived Knowledge

How Max Decides What to Store

Not everything enters long-term memory. User messages scroll through a rolling history window and vanish after ~10-15 cycles. Only items the agent explicitly chooses to REMEMBER are embedded in LTM. PIN holds temporary task state. The decision gate asks: is this a reusable skill, a proven belief, a relationship fact, or a hard-won lesson? If yes, remember. If it is transient task chatter, let it scroll away.

The Reasoning Feedback Loop

Query retrieves stored beliefs. NAL/PLN reasoning derives new knowledge from retrieved premises. Derived conclusions are themselves stored via remember, closing an autonomous learning loop: memory feeds reasoning feeds memory.