Dietrich Dörner

okay, how does this apply to cognitive architecture? Reinforced retro.

In advanced autonomous agents such as Dietrich Doerner's PSI and micro-PSI frameworks, the system records its experiences as a sequential protocol chain, a temporal record of situations and actions stretching from the starting point to the ending point.

The trigger, when a motivationally relevant event occurs at T sub F, the end, such as an internal demand or urge is suddenly satisfied or heavily violated, the procedure reinforced retro, meaning retrograde reinforcement is called the backward sweep.

The system doesn't wait for a future cycle to learn.