What should we forget? A computational model of memory consolidation
This work provides a foundational computational principle for memory consolidation, relevant for understanding how biological and artificial systems efficiently learn and adapt from experience.
This paper proposes a computational model of memory consolidation, termed scaffold-flow memory, which addresses the problem of retaining relevant distinctions while discarding irrelevant variations from past experiences. It demonstrates that an intermediate granularity of representation minimizes future risk, as overly fine representations waste capacity and generalize poorly, and overly coarse ones merge situations requiring different responses.
Neural and immune memory rely on different biological mechanisms but face the same computational problem: future situations rarely repeat past ones exactly. Memory must retain distinctions that alter future responses while discarding irrelevant variation. We formulate this problem as \emph{scaffold-flow memory}: fast, state-dependent responses constitute the flow, whereas slowly changing physical variables form a scaffold that constrains future dynamics. Consolidation writes a predictive coarse-graining of experience into that scaffold. A useful coarse-graining must preserve future-relevant distinctions, generalize to novel experiences, support approximately autonomous coarse dynamics, and provide enough future benefit to justify its physical cost. We quantify failures of coarse autonomy through a leakage measure, relate leakage to excess future error, and identify persistent memory classes with slow dynamical modes. We further show that when experience does not self-average, storage is necessary rather than efficient. In a Willshaw associative memory, a metastable neural attractor, and a stochastic affinity-maturation model, future risk is minimized at an intermediate granularity: overly fine representations waste capacity and generalize poorly, whereas overly coarse ones merge situations requiring different responses. These results support a common computational principle: \emph{memory consolidation selects a predictive, dynamically usable, and affordable representation of the past}.