From the moment of encoding, two processes run simultaneously: acquisition and erasure. The erasure is not passive decay — it is driven by dopamine neurons that fire chronically, activating Rac1 → cofilin → actin remodeling → synapse shrinkage. Both processes begin at encoding. The race starts immediately.
Sleep enhances memory not only by running the consolidation procedure (slow oscillations → spindles → ripples; see sim 48) but also by silencing the forgetting-cell signal. The same window that runs the NREM procedure also removes the active erasure pressure.
Inhibiting Rac1 extends short-term memory duration. It does not convert short-term memory to long-term. Duration and stability are separate properties of a memory.
Drag the gray sleep window to see how timing changes what survives.
Forgetting cells: Dopaminergic neurons (DAL neurons in Drosophila mushroom body; analogous circuits in mammals) fire tonically and drive Rac1-mediated erasure at baseline. They are not activated by fear or punishment — they run chronically. Memory is the exception that maintains itself against the default output. The Berry et al. (2012) study found that genetic silencing of these neurons extends short-term memory lifetime from ~3 hours to several days.
Duration vs stability: When Rac1 is inhibited, short-term memory persists longer. But the longer-lasting trace is not a more stable trace — it is the same transient trace moving more slowly toward zero. Increasing duration does not increase the probability of consolidation; it increases the window during which consolidation remains possible. If no strong consolidation event occurs during the extended window, the trace eventually decays regardless. This is visible in the STM-only view: both curves reach zero; one just takes longer.
Sleep and the race: The simulation shows why sleep timing matters differently for the two conditions. When sleep comes early (drag the window left), both conditions produce similar LTM — because the STM trace is still near 1.0 when consolidation begins. When sleep comes late, the normal condition has nearly exhausted its STM supply, limiting what consolidation can use. The inhibited condition still has more STM remaining, so late sleep benefits it more. The asymmetry isn't because inhibition makes memories better — it's because inhibition preserves the raw material longer.
What this simulation can't show: the molecular distinction between "extended STM trace" and "consolidated LTM" is not accessible during recall — both appear as memory. The difference only becomes visible over timescales longer than 24 hours, when the extended trace eventually decays while consolidated LTM doesn't. The blank left by a trace that lasted three days before decaying is indistinguishable from the blank left by one that lasted three hours.